Harpreet: [00:00:09] What's up, everybody, welcome, welcome to the artists of Data Science, happy hour. It is Friday, December 3rd, Friday, December 3rd. That means it is the last month of 2021. Dan, the CEO of Buy Quick Spin They have been a crazy year man in 04. For a lot of us, you know, good stuff happening for a lot of us and just all around madness for the rest of the world. Hopefully, you guys got a chance to tune into the podcast, or at least today with Christian Espinosa. He wrote the book The Smartest Person in the Room.

Harpreet: [00:00:39] It was an epic book.

Harpreet: [00:00:40] I absolutely loved that book and I love chatting with him. I learned a lot. That conversation, hopefully you guys get a chance to tune in sometime this weekend to listen to that. Earlier this week, actually earlier this week, yesterday I was on a podcast called How did you learn that? How did you learn that with Andrew Berry? That's his name. The Curious Lion That was a dope podcast. I really, really enjoyed being on that podcast. Shout out to Gilbert Eichenbaum for for setting me up with that. Interestingly enough, I thought this was really cool. So. He had actually interviewed somebody who's had a tremendous impact on my life, just a few, you know, episodes before before interviewing me. He interviewed Akira the Dawn. I can't believe that I was on. I was interviewed by somebody who had interviewed me. Care of the Don. I care. The Don, if you guys don't know, is probably my absolute favorite artist.

Harpreet: [00:01:32] He created something called meaning wave.

Harpreet: [00:01:34] And so just to to just that synchronicity, I thought I was insane. Hopefully, I can get a cure onto onto my podcast at some point that would be epic activities. I mean, if it wasn't for a cure that Don wouldn't have heard of,

Harpreet: [00:01:48] You know, a lot of

Harpreet: [00:01:49] Things, trust me with @ArtistsOfData Science podcast is not be what it would be without his influence. So that was pretty cool. Coming up next week, man, a lot of good events happening next week. So [00:02:00] if you have not already, be sure to register for the Comet Roundtable that we have going on on the 9th on Thursday, December 9th. There's a link on my LinkedIn profile. If you all you guys watch my LinkedIn, go here, find that event on the calendar and check it out. We're going to be talking to pretty much like a year in review for machine learning in twenty twenty one. We're talking

Harpreet: [00:02:22] About the

Harpreet: [00:02:23] Challenges and surprising discoveries of deploying machine learning at enterprise

Harpreet: [00:02:29] Scale. We've got some

Harpreet: [00:02:30] Awesome people that are going to be part of that panel. We've got Chris Brosnan from The Real Real. He's going to be on art COVID stroke from work fusion all shake kit, who is from Uber, Uber and also the man himself, Mr. Gideon Mendel's, CEO of Ml, will be there as well. I'll be emceeing, hosting, moderating that session, so that will be a lot of fun. Be sure to register for that. Check the link right there on my LinkedIn profile. You'll see it. Also next week, I've got got a bunch of cool interviews that will be going live on LinkedIn for going live with Nik

Harpreet: [00:03:05] Singh, who wrote the

Harpreet: [00:03:07] Book based on the science interview, so we'll be talking with him. And then also that same day, just like an hour later, we'll be going going live twice. I'm going to say so keep an eye out for that. Going live with the gentleman from one salting. So Jerry

Harpreet: [00:03:20] Lee and

Harpreet: [00:03:24] Mr. Javier warm up LinkedIn his name. Mr. Harvey, hear from from one sultan. That'll be that'll be a great conversation. I'm excited to talk to them. Jonathan Javier.

Harpreet: [00:03:34] It, dude. I'd be slipping

Harpreet: [00:03:36] Sometimes.

Harpreet: [00:03:37] Jonathan.

Harpreet: [00:03:38] Sorry about that. If I forgot your first name for a second. But yeah, that'll be great. It'll be a good, good conversation. Hopefully guys can can tune in for that live. Feel free to join on the live stream and ask

Harpreet: [00:03:49] Questions and join

Harpreet: [00:03:51] In on the bigger activities. Hopefully you guys had a amazing Thanksgiving holiday.

Harpreet: [00:03:56] I hope it was a great, great

Harpreet: [00:03:59] Time for [00:04:00] you guys. Hopefully you guys got a chance to capitalize on Black Friday deals.

Harpreet: [00:04:03] I, for one, had bought

Harpreet: [00:04:05] My son a brand new car seat, so we are going to be rolling in style in here. The brand new Greco. So let's let's get let's you know, let's talk about it. It's the first Friday. I mean, first Friday of December, of the last month, first Friday of the last month of the year. That's crazy. You buy it very, very quick. You know, I think it's time that we should celebrate our failures, you know, whether they turned out good or bad. So I'm

Harpreet: [00:04:30] Wondering what what

Harpreet: [00:04:31] Was something if you reflect back on on January 20?

Harpreet: [00:04:34] Twenty one, right?

Harpreet: [00:04:36] A mere 10, 11 months ago,

Harpreet: [00:04:39] 11 months ago, right?

Harpreet: [00:04:40] My math is right. What was something that you had planned or hoped to accomplish by this time of twenty twenty one, but did not come to fruition? So what's something that you have planned to accomplish, but just it didn't

Harpreet: [00:04:57] Happen, whether you tried

Harpreet: [00:04:59] Or didn't try, you know, so it's something that that didn't happen for you guys that you had hoped would happen. Let's start with the Christian, and then from Christian will go to a we'll go to Antonio and then

Harpreet: [00:05:09] Eric and whoever

Harpreet: [00:05:11] Else wants to jump in. Please let me know. And by the way, everybody tuned in on LinkedIn. If you guys have questions, please do let me know in the comment section right there on LinkedIn. If you watch on YouTube, I got AIs on YouTube as well, so feel free to ask your questions on YouTube and the chat and you are more than welcome to join us live in the room as well.

Harpreet: [00:05:30] Go for a Christian.

Speaker3: [00:05:33] Yeah, I was thinking about this as you're you're asking that question, and it's crazy because this was for me, especially as a crazy year, and I think a lot of people, maybe twenty twenty one has seemed like multiple

Harpreet: [00:05:44] Years in one because

Speaker3: [00:05:46] Of how fast a lot of things have changed this year. But the one thing that I definitely failed at was my my book goal, how many books I was going to read this year, mainly because I don't commute anymore. So forget Audible. I haven't made time for it as much as I [00:06:00] should, so I think I wanted to do the whole like at least a book a month. I think maybe I hit four all year, so it's like a multi-year low for me. I'm going to see if I can change that next year.

Harpreet: [00:06:11] Of the four that you did read, which ones, which ones, which ones were they?

Speaker3: [00:06:16] Continuous discovery habits, which I've referenced before, but Theresa Torres, which was

Harpreet: [00:06:20] Like,

Speaker3: [00:06:21] I won't say life changing, but career changing, I've got this

Harpreet: [00:06:25] One here.

Speaker3: [00:06:26] The product book and this one, I can't really count this. This will be my fifth of the year. I've just started this one that was recommended by John Cutler from Amplitude, which is about trustworthy, online controlled experiments. So it's all been kind of work related, but because I'm kind of in a new field, in a new area, for me, it's been super beneficial and I've grown a lot through them. So I love taking other people's great ideas and not having to reinvent the wheel because most of the time they know better than I do in any way.

Harpreet: [00:06:55] Absolutely. What was that? That that phrase a smart man learns from his mistakes, a wise man learns from the mistakes of others, and it's time you can't. Yep, exactly. So, yeah, so for those you guys just joining in question is, you know, we're celebrating failures, man, why not? I've got a ton of them this year, so let's just celebrate the failures. Let's talk about what it was that you hope to accomplish at the beginning of Twenty Twenty One, and now that you look back, it's December of twenty twenty one and you did not accomplish that thing. What is that thing? Eric Simms, please tell us.

Speaker3: [00:07:29] Two words side projects. Like any good data science professional, you start out whatever, whatever year, week, month, morning with an unrealistic number of projects and ideas. One in particular comes to mind that it's, you know, I mean, it's half done, and I learned some good stuff from it, but I just didn't ever so OK. Give a little bit more context. So I wanted to basically redo a project that I had done all excel based [00:08:00] in a previous job, and I want to do it in Python and then take it and turn it into just like a little bit of an application that could stand alone. And I got through the piece I was really interested in and converting the, I guess, the nuts and bolts of the optimizer into Python. And then at that point, I was like, I hit, you know, you can like, learn stuff that's a little harder than what you know, but this was like way too much harder than what I knew and I was in school at the time. So it was like, You know what, I'm just going to keep learning all this school

Harpreet: [00:08:28] Stuff, and maybe someday I'll

Speaker3: [00:08:29] Come back to that side project. It's not on my twenty twenty two docket yet, but you never know. Maybe it'll find its way out there. So, yeah.

Harpreet: [00:08:36] Side projects like Let's go to Antonio Antonio. Go for it. And then, by the way, shout out to Russell Russell. If you want to go next, please do feel free to go next after Antonio. Everybody else in the room. What's going on? Costa? Gina. Matt Blaze. Monica Croyle was in the building, but she has left the building. Hopefully she does return, by the way. You guys listening in on LinkedIn YouTube, if you guys got questions, you guys got comments. Please do feel free to let me know what they are. But but Antonio, go for it.

Speaker3: [00:09:07] All right, well, one I want to say good to be here. I saw you shouting me out. I was on LinkedIn previously with a baby in my hands. I really didn't join too much to the conversation, but I was still listening when I can. But this year, I mean, the one thing that came to my mind, I had signed up to this no code community, 100 days of no code. And it's a great community. But it was like I was supposed to do this thing for 100 days challenge. And it just doesn't go as planned. I think I made it to like 17 or 18. But then, you know, life happens. We ended up buying a house, my wife getting pregnant and like afterwards, you know, one thing after the other. And just and like [00:10:00] Eric is saying, I mean, it's just there's so many interesting things out there

Harpreet: [00:10:04] That you want to accomplish everything.

Speaker3: [00:10:06] I was like, Yeah, I'm going to do data science projects. And then the reason why I wanted to learn no code is so that I can turn it into applications. The data science side projects didn't work out, let alone turning it into an application part. But hey, at least one of those where if you set enough goals and if you shoot for the Moon and you land among the stars or something, at least you've done something. So, you know, still, some things ended up getting a new job and things like that. And it's always tough because you're like, Should I work on my side projects or try to learn whatever I need to learn for this new job or something, you know? But I mean, at the end of the day, all those things are excuses. I guess I should have done those things. But that's what New Year's resolution is for you. Set things up and you do them afterwards.

Harpreet: [00:10:56] Yeah, congratulations again on the baby man. Can't wait to see some pictures, man. I feel free to it. I mean, yeah, baby pictures. And you know, if you got any questions or anything on fatherhood along the way, do my best answer. Just get a get a snoo. You need a new.

Speaker3: [00:11:13] I also pay your affiliate marketing because you keep advertising now, dude, that's

Harpreet: [00:11:18] It's actually a good point.

Speaker3: [00:11:19] I should holler at them and you should look in.

Harpreet: [00:11:23] Let's go to Russell and then Gina. I think Gina had the hand up. So if anybody else wants to jump in here on on, you know, we're talking about celebrating failures. What did you fail to accomplish that you hoped you would accomplish at the beginning of the year? Because you know, it's OK, it's OK to fail? Russell, go for it. And then after Russell will go to Gina, and I think co-sleep had his hand up. So then after Gina, go to coach, then go for it, Russell.

Speaker3: [00:11:48] Hello, everybody. So first, a quick comment on Antonios update there. So whilst there might have been some failures there, you know you've got a new job and you've got a house and your family is going [00:12:00] well. Those those things kind of blow any other families out of the water. Those are significant things for my friends, so I feel good about it. Russell, thank you. So one of my biggest things to do at the start of the year, a little more obtuse than Data specific was maintain my sanity through the through the year of 2021. You know, kind of the second year following the impact in Kogut and what's, I don't think I've failed.

Harpreet: [00:12:33] It's been more

Speaker3: [00:12:33] Challenging than I expected. I think I was being perhaps a little glib at the outset thinking, you know, we've had 2020, 2021, you know, give it the first quarter and we'll be back on track. And, you know, it didn't happen that way.

Harpreet: [00:12:49] You know, so you

Speaker3: [00:12:49] Know, I'm still here. I'm moving on positively. Like, I'm sure everybody else is, but I feel a little more burdened by the complexities of the last two years than I'd expected to at the start of the year.

Harpreet: [00:13:03] So, yeah,

Speaker3: [00:13:05] I don't think I've been, as shall we say. Resolute throughout the year, not not terribly so, but, you know, it's affected me,

Harpreet: [00:13:15] So I'm mindful

Speaker3: [00:13:17] Of the fact that it is affecting me trying to respond to

Harpreet: [00:13:19] That positively

Speaker3: [00:13:21] And and move forward and take advantage of all of the improvements as they come. Although the irony of this new.

Harpreet: [00:13:31] That is a is

Speaker3: [00:13:34] Is not great time, but still hopefully with the way things are going throughout the world

Harpreet: [00:13:39] And

Speaker3: [00:13:40] Send safety measures, this is going to be a short lived thing and we're not going to go through another year of of impacts from this. So beside that Data Data, science wise, I wanted to expand my network, which I have done. You know, I have not expanded it, but perhaps haven't expanded [00:14:00] it quite as much as I'd like to. I did want to go to some of the events around about. I didn't go to a single physical event, but all through the year. And whilst I did go to some virtual events, there's been virtual events that I failed to go to as well, so haven't been nearly as active as I'd like to

Harpreet: [00:14:20] Have been in the wider community

Speaker3: [00:14:23] So that I'd say that's probably been my my biggest failure over the last year.

Harpreet: [00:14:27] And I think he got the right mindset going into it. Russell, thank you so much for sharing that. Shout out Ted Barrett just joined Mexico. Wassup Joe Mama Joe is in the building. Thank you so much for joining, so we just kick it off the top of the hour just talking about, you know, just celebrating failures. What's something that you wanted to accomplish at the beginning of the year but just failed to accomplish? You know, now that it's December, just you looking back like, fuck, and I got that thing done, but I did not. Let's go to Gina. And then by the way, anybody that got questions, feel free to let me know I'm going to add you to the queue. Or if you if you got questions on LinkedIn, go ahead. Post your question on LinkedIn comment section. I'll get to it as well. Gina, go for it.

Speaker4: [00:15:05] And hopefully you can hear me. Yeah, OK, cool. I'm off camera because I'm out and about moving around, but I wanted to try to celebrate our failures. My failure was not getting a job and. You know, I guess well, for me, I'm a career changer, and although I've done a lot of work, so it's case that this is so, so much different Data science, but of course, the recruiting process. And the function is different than what I had done before, and I'm glad Russell

Harpreet: [00:15:44] Talked about this

Speaker4: [00:15:45] Past year and the mental health aspects, and I think for me, in addition to the shock of, I mean, the pandemic has been terrible. But the shock

Harpreet: [00:15:58] Of all the political upheaval

Speaker4: [00:16:00] Here [00:16:00] in the states, it's just, he said, the past year kind of had a bigger impact on him than he expected, and I feel the same way. So, you know, it's interesting. I think, you know, we think we can weather these things. I do think there are some kind of systemic changes going on in societies.

Harpreet: [00:16:21] And I think central

Speaker4: [00:16:23] To it, frankly, are the algorithms and data science. I really do feel like this is a challenge. Maybe one of the biggest challenges of our time is how to deal with. Misinformation, how to help people understand what they're looking at, how they're. Other chains being yanked, frankly, because I think that's a lot of what's going on in the world, those chains are getting yanked. And unless we which is hard, unless we engage our cerebral cortex more and say, OK, hold on, wait a minute, we need to, you know, what am I reading? Is this real? You know, I just kind of get amped because it's, you know, it's triggering the lowest common denominator emotion. All these things, I mean, they really have weighed on me because prior to the insurrection, frankly and does everything. I mean, I was worried about it. But now it's like, I mean, honestly, you know, I think democracies. Many democracies in the world are in trouble, and it's because bad actors use these tools in some cases, and it's complicated but manipulating people in ways that I think a lot of us are really aware of. Even if people talk about it, they kind of know. But then it's hard because we still react to stuff. So the failure of getting a job is complicated, of course, and it's depressing. It's [00:18:00] hard to celebrate that. But I'm just trying to look at

Harpreet: [00:18:03] It as well. You know, there's

Speaker4: [00:18:05] Been so much change. I've got this new tool kit to add on to my existing experiences, and I really, you know, it's just a matter of, well, how do I navigate? How do I use them to best effect? You know, do I go more into the environmental sustainability realm as I thought I might? Or do I go and just get a job and really bone up on the the skills and a in a rural setting? Or do I pursue this other angle which seems far more nebulous a welcome? Any comments that anybody care?

Harpreet: [00:18:44] It means so much for sharing. I appreciate that, and to the point about talking about, you know, kind of responsible Data science, there are a couple of weeks on December 15th I'll be live streaming with Grant Fleming. He wrote the book Responsible Data Science.

Harpreet: [00:18:59] We'll be chatting live. So be sure to to

Harpreet: [00:19:02] Keep it

Harpreet: [00:19:03] Out for that invite and

Harpreet: [00:19:04] Join if possible. And if you guys have any particular questions that you would like to have asked of Grant, please do let me know.

Harpreet: [00:19:11] I'll try to,

Harpreet: [00:19:12] You know, incorporate that into the

Harpreet: [00:19:14] Into our

Harpreet: [00:19:14] Dialog. So shout out to everybody else that just joined in, well, just one of the person that recently joined in. Mark, what's going on? So question we're we're you know, I'm just kicking off the discussion right

Harpreet: [00:19:24] Now,

Harpreet: [00:19:25] But anybody can take the discussion anywhere they want. I was just asking, what are some things that you wanted to accomplish at the beginning of the year, but you did not yet accomplish? If anybody wants to chime in here, that just joined. Raise your hand. Let me know and I'll I'll call you on to the stage, so to speak. So does not. Yes, Eric. Go for it.

Speaker3: [00:19:48] Ok, so I just thought of another thing that I wanted to accomplish. I think it was at the beginning of the year end of last year, whatever. So but I didn't accomplish it and I'm actually not worried about it. So. [00:20:00] I had the opportunity during my classes and stuff to do some different certifications, right? And like, I was super busy with school and I made time to do some of the certifications, but I did not make time to do all of the certifications and and like in the moment when you're like feeling like everybody has all the certifications and you're dumb, then it feels like all the certifications are way important. And like all the certifications are not important. You know, like I won certification in particular I did not get was my

Harpreet: [00:20:38] Awu's whatever the noob level certificate is

Speaker3: [00:20:41] Like, I don't have that. I would like to learn some of that stuff eventually, but I'm not kicking myself because I don't at the moment. I spent my time doing other stuff that was valuable, and so I think I would just put that out there to anybody who's feeling like, Ah, should I get this other shirt? Like, maybe, maybe not like, don't kill yourself over it.

Harpreet: [00:21:02] I have precisely zero

Harpreet: [00:21:03] Certifications, so

Harpreet: [00:21:05] That makes anyone feel better. I've got I don't even have a certificate to podcast. I'm still like, You're doing it, Eric. You also wanted to shout out Roy Gibler, man Roy. I haven't seen Roy in such a long time Roy. Give little ray ray. Ray Rice. Hey, Roy Ray Ray Gubler. Yeah, he was like he was a regular fixture at the officers around this time last year, but I think he's since gotten busy during this time. So, Ray, if you're listening by any chance, I'd love to have you back. But yeah, talk to us about a little. First of all, he does drop awesome tableau knowledge like amazing, amazing tableau knowledge. Talk to us about the tip that he gave you that helped, and then after that will go to the coast of Denmark.

Speaker3: [00:21:44] Yeah. So like the cool thing. So I don't rate does like Tableau stuff all day, every day. And he

Harpreet: [00:21:49] Has a ton of tableau like

Speaker3: [00:21:51] Tips and stuff that he's written down for himself. And so he shares on me now. And a few days ago, he shared a tip

Harpreet: [00:21:57] About

Speaker3: [00:21:59] Using captions [00:22:00] on your worksheets to leave like instructions or notes and things like that. Because and I had just been thinking like, last week, how can I do this better so that the next person who comes along can read it without cluttering up the whole dashboard with instructions? And so he said, Hey, do this thing with captions. And I got stuck. I had to leave what I was doing in the middle of creating like a few different fields or all reliant on each other. And there's no way I would have remembered, even like the next day what I was working on. And so I just took the time, wrote it all out. It was right there. I didn't have to dig through my notebook or anything to find it. And then today I sat back down, worked on it, got it finished. And it actually, fortunately was part of something that will make a big difference for one of the products that I support. And so extra good vibes from that. So if you don't follow regular follow regular for a cool tableau tips. Plus it's just a nice guy.

Harpreet: [00:22:53] He got some good knowledge like like, I haven't seen his stuff show up on my feet for like ever. So I'm like going through his thing right now, just smashing like on everything. I feel like my feet has been nothing but sponsored ads lately.

Harpreet: [00:23:03] I don't understand why

Harpreet: [00:23:04] That is

Harpreet: [00:23:05] Off for a

Speaker3: [00:23:06] While, but he's back. He's been back for a few weeks.

Harpreet: [00:23:08] Right now, we miss you. It's been a while since you have been here, so I miss having you around. So if you're

Harpreet: [00:23:12] Listening, come back and

Harpreet: [00:23:14] Hang out with us some time. Let's go to Kosta and then after costar mark.

Harpreet: [00:23:19] So I guess

Speaker5: [00:23:21] A couple of things came to mind and stuff that I haven't succeeded in. First thing is my personal fitness goals, right? I feel like we talk so much about the technical growth and the other side of life that often what gets hit is like. And I felt that it's gone through kind of waves, right? It was going great at the start of the year. Then in Sydney, we got hit by the Delta variant, got chucked into another lockdown for three three months or so, and that dropped it down further. And it's just been up and down. And the same thing kind of. I noticed a similar pattern with my learning, with my technical learning, right? I do it in these bursts, in waves and it's like, you know,

Harpreet: [00:23:55] I earlier, just what,

Speaker5: [00:23:57] Eight months ago, nine months ago, I was the the [00:24:00] GTC summit at the start of the year doing courses on, you know, and video stuff. And then now it's like, now I'm learning all about GCP on Coursera, and I've got like a stack of textbooks here that I just haven't

Harpreet: [00:24:11] Got around to because I've been working

Speaker5: [00:24:12] In like this burst mode, right? Right. Come out of the woodwork saying, Okay, cool, I need to get on it right. And then I spent two weeks just smashing stuff out and then maybe a month where it goes great and then another month, which is kind of, you know, it all dips down.

Harpreet: [00:24:26] And what I'm kind of realizing

Speaker5: [00:24:28] Based off what Russell said, what Gina said and what a few others have said is a lot of it comes down to resilience and change,

Harpreet: [00:24:36] Right?

Speaker5: [00:24:36] And focus is, can I? So it's almost my goal for 2022 is to start looking at can I retain my focus irrespective of what changes like, we haven't seen more dramatic changes in the last 30 years than we have in the last two years, at least widespread across the

Harpreet: [00:24:54] World, right?

Speaker5: [00:24:55] There's definitely been major pockets of it, but widespread across the world. We haven't seen this kind of dramatic lifestyle changes that affects literally everybody. So it's a bit of a learning curve on how to do that. And it's

Harpreet: [00:25:06] Only kind of in this conversation

Speaker5: [00:25:07] That I'm like, You know what, this all of these other things, I haven't hit my goals on. On studying certain textbooks worth of work or I haven't hit my goals and certain fitness aspects, and it's predominantly because I've lost that focus being distracted by change. Right now, I've got no clue how to solve that, as it's a focus thing. It's a mentality thing. It's yeah, I think the first step in solving that is identifying that hang on. That's that's the main problem, right?

Harpreet: [00:25:39] Thanks so much for sharing Coke step. If anybody wants to respond to

Harpreet: [00:25:43] That, please feel free to do

Harpreet: [00:25:44] So after we get to to mark. Mark, go for it.

Speaker3: [00:25:49] I think I think the one thing the gold I'm truly trying to get down is my prioritization skills and being able to stick to those part. The discipline, two of those privatizations.

Harpreet: [00:26:00] So [00:26:00] I feel like I

Speaker3: [00:26:00] Got really down in my in my day job. But like outside of my day job, I'm like, all over the place definitely fallen for that shiny, shiny thing since jump. Jim Webb three is my my latest distraction, and I'm actually kind of Harp. I'm actually probably prioritizing more, actually, because I was falling in love with it. But like with me starting to kind of like my side hustle and it being like a full business and everything really clicked in my head, like how much of my lack of prioritization is really impacting me. And so I've been slowly changing my behaviors to to make it right. So, you know, going to bed at the right time, eating right, all those different small things allow me to do more and prioritize accordingly and be able to say no to a lot of things. Super important that I struggle with in my personal life, but I'm slowly getting better and being forced to get better because there's a lot of cool opportunities coming my way through, through my side hustle that requires me to be way more focused and disciplined. And so I'm looking forward to next year building on successes I've had with that. Am I in my job to bring that to other aspects of my life? So I'm pretty, pretty excited about that. But good

Harpreet: [00:27:16] Direction.

Speaker3: [00:27:17] Just you just got to execute.

Harpreet: [00:27:20] I can I can relate to that thing about just I don't know, it's just I'm somebody who just has I'm just naturally curious. I have a ton of interest and. They meander like I study everything from physics to philosophy to Web3 to various different programing languages, and just all this stuff is just super interesting to me. And it does kind of, you know, I guess you're saying shiny object kind of syndrome. A little bit, but but I follow that thing until my curiosity is exhausted, then move on to the next thing, but then try to find links between what I was just

Harpreet: [00:27:54] Studying or what I was just

Harpreet: [00:27:56] Reading about and whatever to what I'm currently doing. But yeah, Web three [00:28:00] is definitely the newest fascination for me as well. It just. There's rare opportunities in life where somebody can be on the frontier of something and be part of a movement in its infancy, right? Like I fucked up when I was younger and I couldn't be on that wave when when you know, the Web 2.0 is happening, even though I could have right, I just live choices. And now I feel like this is a chance to like, kind of redeem myself with that and some using that as a opportunity for me to just to get on the ground floor of something that's interesting and and capitalize on that. But Charlie Dow is where it's at. If you guys do not know what Charlie Dow is all about. You're looking at three members of the Dow right here. We got Mark Makiko and myself. I'm looking forward to chatting with. With Carlos next week.

Speaker3: [00:28:54] Hey, I'm part of it, too.

Harpreet: [00:28:55] You're on the Dow, too. Damn. All right.

Speaker3: [00:28:57] Well, I don't want to miss out on the shiny billionaires. I don't want to be the only poor one.

Harpreet: [00:29:06] What are you saying, Joe?

Speaker6: [00:29:07] I said he bought himself a gaming PC, though he might be out of commission for a bit.

Speaker3: [00:29:12] I'm going to be staking those altcoins. You know, that's why the graphics on this.

Speaker6: [00:29:17] All right. I'll give you some perspective on this. I'm a lot older than I look. I got on the internet back in ninety two before there was even a web browser. And so I think to your point, right, you feel like you're on the the cusp of whatever New Frontier is. I went through a few of these now, and it's interesting like the the trends that I've seen over the years. It's like there's the there's the stuff that seems obvious, right? And there's what I found is it's the obvious stuff that usually is what takes hold. So I always I would always keep a lookout for those things to the obvious stuff is what everyone's trying to do right now. And I think that's great to get into. But keep an eye out for the applications [00:30:00] where it's not intuitively obvious, because that's where the I think the real opportunities are made is sort of the stuff from left field right, even Facebook. That was a left field thing. It seems the most obvious thing in the world now, but trust me, when they came out, it's like, this is like the dumbest thing I've ever heard of. Why would you ever do this? Twitter was the same way. So you talking about Web 2.0, Social Web 2.0? It seemed like a terrible idea, actually at the time. So I remember when that came out and I was like, well, user generated content, because obviously that's a great idea. Well, look at it now, you know? And so decentralizing everything, I think is the just the way forward. But the thing I would caveat that with and as you guys get into, you know, Web three, which I think is amazing, like just keep in mind all the left field applications that nobody's really paying attention to because a crazy or something sounds it. Actually, that's the stuff that makes the most sense.

Harpreet: [00:30:48] So it's this thing about the stuff like come out of left field, it's and just remixing that idea with the. The curiosity, right, like just being curious about a number of different things is you can find those intersections between between things, right? And find something interesting in that little niche.

Harpreet: [00:31:08] You know what I mean? And that's kind

Harpreet: [00:31:10] Of where you find those left field type of opportunities. I mean, I was on I was on Friendster, man, like, really, really date myself, man, Akiko said.

Speaker6: [00:31:22] She was too at the I was on the internet. That's all. So I'll just shut up now. All right. Fun.

Harpreet: [00:31:27] You were two in 1992. Damn, I feel old, man. Shit, I didn't.

Speaker7: [00:31:31] Well, that's the that's the beautiful part, right? It's like I look left and right and it's like, you can't tell. Like in the Data science machine learning space. I'm like, I don't think you can kind of tell someone's age necessarily. You can kind of look at like maturity and experience, but even that's not it's not like a super good indicator. You know, so I don't know. I think it's funny just because like I'm on a team of like all women, you know, email ops [00:32:00] like engineers, which already is a little bit just unique. But also we all happen to be within like one or two years, like years of age. But like either the other two ladies are just way more experienced in like engineering. And then I have just a lot more experience like the Star World. So it's like it's a little bit weird, right? But you get a you get a nice team together with different skills and experiences, it eventually blends. But yeah, it's just it's so hard to tell because I

Speaker6: [00:32:26] Think that's really cool about the tech space and the tech technology in general is, you know, I think for as much of a lack of diversity there is in this field, it's a meritocracy, unlike a lot of other fields or there's gatekeeping like you trying to be a doctor with the same premise as being a software engineer, right? No, it's impossible. You go, you'll be in law. And I think rightfully so, actually, in that case, because I don't want anybody just being a doctor, that'd be horrible. But like a lawyer or that kind of thing, right? I mean, or getting into an exclusive company, you know, being a business person, it's like tech is one of those things where I think it's really rare in the sense where you it really depends on your skills. And if you're keeping that current, if you aren't like, you know, you're a dinosaur and there's a two year expiration, Data notice is almost everything. It's a two years is kind of where you got to start recycling, so it's cool.

Speaker7: [00:33:14] I think it's also good when you get people from, I mean, totally biased on this one. I think it's good when you get people from other fields, too, because I feel like if you're if you've been in like engineering too long, like people adopt a lot of like weird mindsets. Like I had a person who they were interested in applying to my team and they're like, Well, I'm a test engineer. I don't know if I would fit in with, like the AML ops team. You know what skills and things do I need? It's like, Well, you've got Python and Docker and. Kubernetes is on your resume, so I feel like already that's if you're just focusing on infra and platform already, that's like a really good spot to be. And everything else is just like, can you deal with like JSON smell, [00:34:00] whatever files? Can you learn quickly? Do you know what cloud is like? It's just more generic stuff. And it's kind of like, we sort of figure out the tooling to fit the problem, but it's I don't know. It's kind of interesting, but like I've seen other like gatekeeping sort of attitudes, and I just feel like that's when people have, you know, they've had the Kool-Aid a little bit too long and they're just now like, Well, if you're like a test engineer or if you're a product engineer or if you're a data engineer, then you can't do XYZ.

Speaker7: [00:34:30] It's like, No, no, they just they just figure out what they need to know. And then, that's it, right? Like, shouldn't be a huge issue. Some of that stuff I don't get and I feel like in a way, because I come from outside engineering, I'm like, Well, all of you think I'm stupid anyway and like, not accomplish. So I can just kind of go in and learn and mess stuff up. And you're going to like, think it's because I didn't study engineering anyway in college. So in a way, it doesn't matter because I'm no matter what you're going to think of me less, then I think of me so I can just, you know, but but still, I thought it was really funny because it was just like, I don't know if I could transfer in. I'm a test engineer. I'm like, No, you can totally transfer in. Why not?

Speaker3: [00:35:13] Like, it's all good. You know, more people like that.

Speaker5: [00:35:18] I think that's something to be said about software specifically, right? So the kind of space that I've seen, I have seen more of that gatekeeper mentality, even in the other engineering fields, right? So if you've got a mechanical mechanical engineering and mechanical design, you've got electronics. A lot of it gets a little bit more gatekeeper. And I think one of the big reasons for that is that I mean, personal computers, right? It makes software accessible. You're able to go and learn and develop your skills privately, cheaply and reversibly if you screw something up and software. You know, gut check out a previous comment or a previous tag, and you're kind of alright, right? It's an easy fix. You screw up a mechanical design or a product [00:36:00] that goes out and say it's a health care product, say it's something

Harpreet: [00:36:02] Else, but then it starts

Speaker5: [00:36:03] To hurt people, right? So I think software inherently has had this reverse ability, this distributed kind of knowledge sharing kind of style purely because it's so accessible you need a keyboard and a screen, really, these AIs you need the internet, maybe cloud connection, but most of the time you can do it just off a regular AIs computer. So I think part of that really plays into this ability to be a meritocracy. Well, in the mechanical engineering field, if you need to, if you need to become an expert in aluminum honeycomb welding, you need to be able to afford enough aluminum honeycomb to practice on. And that stuff ain't cheap, right?

Speaker3: [00:36:43] It feels like it's great

Speaker7: [00:36:44] What layer of the ozone layer you're operating at. I feel like if you're like at the top layers, right or not, the top bottom, if you're at the layers that are closest to like the physical hardware space, yeah, it's going to be a little bit trickier. But once you get past like two more like the application or, you know, like the API or whatever level. Yeah, I feel like that kind of gatekeeping is just totally unnecessary. It's like all the arguments between programing languages, right? Like, it's although I after doing a little bit of solidity, I can say that I don't like it. So but that's that's a preference. I'm not I'm not saying anything else but like, you know, but I feel like still, if you learn like one language or you learn one framework, you do have to do a little bit of translation. Like JavaScript still screws me up. I just look at it. I'm like, OK, I need to take a little bit on this. But once you on the more abstract layers, once you've learned the principles or you have like one thing, the gatekeeping to other areas to me just does not make sense. When you get to hardware, I think that totally makes sense, right? Because that is like partially the difference between like the architecture used for self-driving cars versus like medical devices, FDA regulations, yada yada. But outside of that, it's like, you know, [00:38:00] we don't need to make ourselves feel self-important, know we're all special. We're all special in our own way.

Speaker5: [00:38:07] I think I think part of it, it comes down to that risk and risk and liability thing is how much is is a screw up going to hurt someone and how much is someone going to have to pay to fix the screw up? Right. Software? It's easily forgiven if I mess up a piece of software like roll their eyes. Come on, guys. Get it

Harpreet: [00:38:28] Together.

Speaker5: [00:38:29] Hardware, any kind of hardware. If you're if your phone

Harpreet: [00:38:31] Fails, you're

Speaker5: [00:38:33] More mad than if there's a software glitch that they just fix on a patch because it's easily fixed, right? So I think it's kind of the nature of that.

Speaker7: [00:38:45] For the record, I didn't say solidity is the worst language ever. But I didn't say it's the best line I'm holding, I'm holding my opinion on it, I'm going to learn some more. Also, if anyone's interested in programing languages, the reason this book I've been raving about to everyone, it's called Crafting Interpreters by Robert Nystrom. And if anyone knows, like the Wizard and the Dragon, book Dragon books about compilers and interpreters and all that, I wouldn't know it because I didn't actually study engineering in college, but apparently that's where most people hit those up. But it's like cracking servers by Rev. It actually walks through like, how would one design a language and how would one craft the create the compilers and interpreters that would then execute and like, you know, turn it over to like byte code and all that and also, I think, does a really good job of explaining it. It's just such a good book. I have to rave about it. It's yeah, it's giving me a great appreciation for some languages, which I didn't learn before.

Harpreet: [00:39:45] You can drop a link to that in the chat. That'll be helpful. Be sure to include that into the show notes. That is solidity is solidity, it's cool, it's interesting that just the idea of smart contracts,

Harpreet: [00:39:57] I guess that's

Harpreet: [00:39:57] That's what I absolutely love. And [00:40:00] we got we got a lot of Charlie dialup in the building. Where's Carlos at Charlie Dow? What up? A question or comments, anybody got questions or comments or anything, please let me know, let's continue. This conversation has been going good. Monica Royal is in the building. Am I Tripoli, you here? Is your hair purple?

Speaker7: [00:40:19] It is purple.

Harpreet: [00:40:20] Yes.

Speaker3: [00:40:21] Very nice.

Speaker7: [00:40:23] I go from every color of the rainbow.

Speaker3: [00:40:25] It was previously red, so then it turned

Speaker7: [00:40:28] Pink and now it's purple.

Speaker3: [00:40:29] And so who knows what's next? I had a

Speaker7: [00:40:34] Comment to what I think Michiko reference to earlier I've been in and out was acting crazy, but as far as the gatekeeping.

Speaker3: [00:40:42] So you guys

Speaker7: [00:40:44] Know, like, remember the unicorns, the Data science unicorns back in the day, right? Where your programmer, you're an analyst, your bi person, you can do all of it and that's what you want. What I've been seeing lately is that they don't

Speaker3: [00:40:58] Want that

Speaker7: [00:40:59] Anymore. They're like, No, you have to just be the AI or you have to sit down and create your models and algorithms. Are you guys seeing the same thing?

Speaker5: [00:41:11] I'm seeing if I can respond to that, I'm seeing some of that, particularly as businesses are trying to figure out how to optimally structure a larger teams, right? A lot of this comes down to how do you structure a big team as opposed to how do you structure an individual career? And yeah, it's a challenge, and I think it's still quite an open challenge. The downside is, yes, I'm seeing like a shift to more rigidly marked lines in some cases. But I'm also seeing in some companies there's an understanding that the the organization of the team is not a is not a fence that people can't cross over. I've seen a couple of examples of companies where, yes, they're defining the

Harpreet: [00:41:58] Roles better, but

Speaker5: [00:41:59] They're encouraging [00:42:00] that cross-pollination a little bit better as well. So I think we're starting to see this turn into an evolution in team management and team structure, right where? Yeah, I think that the thinking is evolving in some places, but obviously it's not going to evolve everywhere overnight. So.

Harpreet: [00:42:17] But I get what you mean.

Speaker5: [00:42:18] Yes, it can be scary where you're like, Oh, are they drawing the lines now? And. You're going to be stuck with the same problem we had 20 years ago, right?

Speaker3: [00:42:26] You know, the antenna is going away. That's sad.

Harpreet: [00:42:31] He says getting harder and harder to to do everything. I think there's unicorn. It's not just

Harpreet: [00:42:37] It's just like some

Harpreet: [00:42:38] Mythical idea. I guess it's hard to do everything. It really is. So actually, I forgot about Mark's question mark, had a Kubernetes question, a lot of Mark's questions and Kosta will go to you because Mark AIs was asking about this for a while. Is that is that all right, coach? Or are you speaking on something that was that? Yeah, cool. I call go for it. So I was going to say, if anybody want to respond to Monica's question, do let me know, I just glossed over the, you know, the responses for that. So if he really wants to please, do I respond, Marc, if you want to respond to Monica's question?

Speaker3: [00:43:13] Yeah. So this great article on Kubernetes basically talking support for for Docker and where this came up for me, I'm writing a piece for a client and kind of doing market research, and the vendor I was looking at was heavily in Kubernetes and using Docker, and it posed this question of, well, this this vendor is heavily like the infrastructure is built heavily on Kubernetes and Docker. But if Docker is no longer being supported for other kind of interfaces for the for those images, you know what, what happens to a lot of the companies who are really depend on that combination of services? You know, how [00:44:00] would how would leadership kind of do that? Like, would they start going through a scoping process of like, I will replace this with, you know, would they would they stop using Docker? I don't know. So. So kind of goes beyond the article. It's like, say, for instance, you read that article of like, Hey, Kubernetes is no longer supporting Docker. You know, what would be your next step if you're one of those organizations who are in that in that spot, we're like, Oh wow, we have to rethink how we structured things.

Harpreet: [00:44:30] You do a good job for this one.

Speaker3: [00:44:33] Thanks.

Speaker6: [00:44:35] That's an interesting question. I mean, it's kind of it's been like Kubernetes. I think I actually stopped using it around the time they announced. I think it was like last year around this time or maybe before they had. So they're kind of not deprecating support, but they're definitely moving away from like long term support of Docker and I think moving towards like a containerized runtime interface or something like that. So their own sort of thing. But. So I just see whatever it is Kubernetes wants to use. I mean, it's kind of a dumb answer, but it's it's the answer I got. So. Before I recall, is not

Harpreet: [00:45:11] Really

Speaker6: [00:45:14] It's not really a runtime interface anyway, so he should be able to just repackage your stuff quite easily.

Harpreet: [00:45:18] So yeah,

Speaker6: [00:45:21] I say good luck on that. But Kubernetes is, you know, I guess if you're if you got enough horsepower to use that in production, then you got enough horsepower to figure it out. You know, container and package dependency management. So that is a beast to work with, which is why I haven't been using it lately. So.

Speaker3: [00:45:44] Sure, for sure. This seemed like a huge wrench to a lot of things as as people are like really trying to free up space for all of a sudden, this like kind of this do this typically kind of quote unquote best practice kind of going away?

Speaker6: [00:45:59] Yeah, because for [00:46:00] a long time, it was almost synonymous with containers and Kubernetes, right? And now it's now it ain't. But what I recall, though, Docker also made some interesting decisions with their business model. I think I don't know if it impacted this or not. I could be talking about my rear, but I think that, you know, Docker did some interesting stuff as well that may have influenced this choice, so we'll see. The only constant in this business is things change, so we should have fun.

Harpreet: [00:46:32] Greg, actually, Greg, if you don't mind, let me go to unless you're talking about doctors community.

Harpreet: [00:46:39] Yeah. Ok, great.

Speaker3: [00:46:41] Yeah. I was going to say I was reading something from a friend this morning who was talking about Kubernetes. He was saying something like quinine is becoming more of a we talked a lot about on premise as being the legacy system, and because Kubernetes requires so much intervention by experts and things like that that

Harpreet: [00:47:06] It is made, in fact at some point

Speaker3: [00:47:08] Become. You know, legacy systems and especially with how much it takes to manage, you need professionals and things like that, and you've got to be pushed to leverage serverless infrastructures to perform those things where you don't have to worry about managing that, that heavy workload or. So, you know, maybe you should look into that a little bit.

Speaker6: [00:47:37] I do have something to say on this insider baseball. Don't quote me on this, but I had heard that Kelsey Hightower, you know he is Mr. Kubernetes. Yeah. Anyway, he was the one who was like Mr. Kubernetes over at Google. I hear he's working on cloud run now, which is the not the guy from police [00:48:00] academy, Eric. That was that was it's funny. So no, he's working on a cloud run from what I heard. So that's a serverless,

Harpreet: [00:48:10] Yeah, container thing.

Speaker6: [00:48:11] So I should tell you a lot if that's true, right? Yeah. Because he, you know, he was a spokesperson for this and put his entire career on, you know, Kubernetes and now, you know, but I agree with that assessment. I think that I think that it's just if you've used it in production, it's fine, but it's also just super unwieldy, like it's a beast to work with.

Speaker3: [00:48:31] So, yeah, the the the surprise of surrealist from what I mean, people I've talked to across companies is that it could it could could come at you with that bill that costs right. So it needs and needs to be more work to be done to make it more affordable, but it will get more affordable economies of scale. But to to answer your question

Harpreet: [00:48:52] Mark, I think the next step for

Speaker3: [00:48:54] Things like Kubernetes is probably serverless infrastructure, cloud infrastructure to see where it goes.

Speaker6: [00:49:01] But I think what your question, though Mark has to do with all of the mobile apps frameworks, though, right? And how they have to work with Kubernetes to like cube flow by definition, by its own name has shortened Kubernetes. So it's like, I don't know, I kind of go back and forth in these kind of attaching your your framework too closely to the underlying infrastructure because he saw the same thing with that mobile flow too

Harpreet: [00:49:22] And spark

Speaker6: [00:49:23] And stuff. So I don't know. It's interesting.

Speaker7: [00:49:30] Mickey, go. Yeah, so I guess the way I had interpreted the change was that. So Kubernetes is like container orchestration, but Docker is not the only container containerization technology out there, and frankly, it wasn't even anywhere the first right. There's been variations of it, it just became synonymous. And I think what I read from just from the updates was that essentially like Docker and Kubernetes, even [00:50:00] though people had presumed that they were really tightly coupled, they were still making their own developments in different directions. But whenever Docker has to make changes, Kubernetes would have to kind of respond to it. And so there's like OCI, which is trying to do some kind of standardization, protocol or or standards for containers. And so Kubernetes was like instead of kind of continuing to support the Docker shim and in general, you hate shims. Shims are always like temporary. You just kind of use you use a shim to kind of just make things plug in for temporary time. They were like, Yeah, we're not going to keep doing this because Docker as a container should follow like OCI, like they should, they should try to, like, play nice. And so it's interesting because we at the Data of media, like we have a monthly book club slash podcast club, like at MailChimp, we were last month was talking about Data meshes and a few of the like. A few of the staff and senior engineers had their reading and they're like, OK, so basically the premise is push the problems to the left. I just push the problems on someone else for Data mesh.

Speaker7: [00:51:09] And it kind of just seems like that's like what? What Cubans decided to do with that being said, like cloud run is kind of interesting because there's like a couple of things, right? Like, for example, if you're looking at the ability to do like async processing like that would put cloud run as a potential competitor to like Kafka and pump some. But on the other hand, it also has some other additional stuff where it can handle heavier workloads than cloud function. So it's like in this kind of like interesting area, and it's kind of it's going to be interesting to see sort of like where they sort of run with it, especially when you like when you think of it as like kind of replacing jobs because that's the other part that's like really painful, right? It's like, how do you coordinate your like [00:52:00] Kubernetes and your and your airflow, right? Like, that's just it's such a pain. And I think most people, they're like, Well, if you had something that was both scheduling and orchestration. I was also kind of serverless that would sort of like kill three birds with one stone. You would you would get rid of like basically like things like pub sub and Kafka. You get rid of like air flow and you'd get rid of Kubernetes. I mean, I don't know, I think it's it's an interesting challenge. We'll see how that actually kind of how that runs because no one likes to kind of let go of their infrastructure that they've committed a lot of like sunk cost into, but.

[00:52:42] It'll be interesting.

Harpreet: [00:52:47] I worry that a lot of that went over my head as well. Listeners engineering talk is in depth.

Speaker7: [00:52:53] Well, actually, Joe like, well, Joe, did any of that sound like totally off? Oh, that sounds on.

Harpreet: [00:53:00] So just the entire segment of this engineering talk that learn a lot is going to wrap my head, I'll be listening to that over and over again.

Speaker3: [00:53:10] Oh, don't feel bad because it's over my head too. I just try to learn the high level. Yes, yeah.

Speaker6: [00:53:16] But I think the trend that we're seeing and writing about this a lot, I just the increasing level of abstraction in the stack and the Data stack is just it's happening across the board. I think the big challenge is going to be interoperability between all these various systems, whether it's open source or a third party that's abstracting all this.

Harpreet: [00:53:34] That's a really big issue right now.

Speaker6: [00:53:37] I think you're going to see a big trend towards metadata is sort of being the glue that holds everything together. Expect to see a lot more of that in the space in the next year. But yeah, because it's become obvious that the way people are doing it now is just sort of a a bridge to somewhere else. But nobody, I don't think anyone's going to say, Oh, this is exactly what we should be doing the way we're doing it now. But, you know, it's cool.

Speaker3: [00:54:01] I'm [00:54:00] really I guess, my question. Oh, good.

Harpreet: [00:54:04] Oh no, he said a bridge to somewhere else. I was like, bridge to somewhere else is the metaphors. Go for

Speaker3: [00:54:08] It. Yes, right? As I say, my next question is like just trying to think like, I'm trying to learn more about like, how do I build infrastructure for Data products? And so I originally thought I was going to be Docker and Kubernetes to learn about because I find it interesting. But after seeing this, I'm like, Oh, you know this, this may not be the the best thing to focus on my time. And so like thinking wise, like, you know, if you're if you were to learn this kind

Harpreet: [00:54:42] Of ml ops kind of

Speaker3: [00:54:43] Thing stuff again, you know, and you have limited time, what would you double down on?

Speaker7: [00:54:49] And it services one hundred percent.

Speaker3: [00:54:52] What services, what you mean by that?

Speaker7: [00:54:54] Yeah, so so Google has. Ok. So this is how I would do it for for any cloud vendor, I would definitely choose one. It really doesn't matter. Adb's or GCP. I go with you, Typekit, just because that's what's at work. But SWC also is you can go with that as well. What's nice about and I'll refer to GCP in this, assuming that also has kind of like Mirror Resources. So if you go into the docs like they'll have a lot of use cases, like if you want to have this kind of architecture or whatever, this is how you can use the different services together. That, to me, typically is very useful because it allows me to understand how the pieces kind of work in terms of what I mean by managed services. So for example, Google has like geeky Amazon has a managed version of Kubernetes if you really want to go with it. But I would almost go like, start really simple to then going really complex. So the simplest would be serverless 100 percent. It tends to be a little bit more costly and also [00:56:00] like they will have limitations on the amount of data you can pass. I think when serverless functions first came out, that was like the really big thing that I noticed was that it was just really expensive to pass more than how many megabytes of or even megabytes, 30 bytes or bits of data bytes or whatever.

Speaker7: [00:56:19] But I would go serverless if that is just like cost wise, just doesn't work out, then I would go towards more like seeing if you can kind of like ship containers or something like that or you. Yeah. And you can use like geeky, but also sometimes too, you can just directly, you can build containers and then you can just directly deploy it so that it gets hosted on something like Heroku or like a cloud version of it. And then I would sort of keep going down or up the complexity stack until you get to like the right spot. I wouldn't sort of go out and learn Docker and Kubernetes and like Jenkins from like immediate get go because you can just it can cause so much grief. But at the same time, if cost isn't like a huge barrier, I would just go very, very simple. Go with manage. Go with things that you can just do, like wrapper scripts around and then start kind of going up the complexity stack.

Speaker6: [00:57:18] Mark, I think between what you were saying, I think I also heard you say like, you are looking for things to help you decide how to design systems, right and infrastructure. If I'm not mistaken or you want to develop them.

Speaker3: [00:57:30] Think about that because one of my interests of mine, especially being in a startup that's very top of mind and also like a lot of my writing I do for my clients, is thinking about where these vendors fit within the market and the ML workflow. So it's having a broader understanding. This really helps.

Speaker6: [00:57:45] I think you need to study architecture. Yeah. So actually, I just got this book from O'Reilly Software Architecture. The hard part's just came out a few weeks ago, so this is good. I've been reading this [00:58:00] partly because I like Neil for the ThoughtWorks people. They do a really good job writing about architecture, but I'd say you should study architecture patterns that might be where you need to go. Yeah, software. Yeah, I like that book to this, but it seems like that's if the other good, solid understanding of architecture, of which I will caveat that I don't think any architect has a good, solid understanding of architecture. You as an architect, what architecture is there? That's a good question. I have no idea, actually, but I would say study this stuff too. It's sort of it'll sort of bridge the gap between, I would say, like the kind of the implementation and the engineering pieces and the broader picture of what you're trying to design and kind of where all the pieces fit right. Architecture is about figuring out the trade offs and figure out what pieces you're going to fit, where basis against constraints. So.

Speaker3: [00:58:51] So yeah, that's the main thing in my my writing is just build versus buy argument, like bouncing between that. So that's really helpful. Mark, if you are if you're a visual learner like me. I can point you to something

Harpreet: [00:59:06] That I do, too,

Speaker3: [00:59:07] To go to YouTube. Get on the channel called. This is my architecture. It's based on AWAs, the invite, a company. That bring a use case they have aboard to build the architecture, some of them are male based. Some of them are not. Or regardless, I try to watch one episode per day to kind of understand how to build architectures that are actually in production. So it really lights up my mind in terms of the common systems that keep getting called based on the use cases, email based or not to see how the whole thing works. And they're like five to seven minute videos. I think this is one of the best like

Harpreet: [00:59:52] Channels that I can work

Speaker3: [00:59:54] With because I don't have patience to read a book. And over time, when I read about company use cases, [01:00:00] I typically get a good understanding of how things are pulled because I try to see.

Harpreet: [01:00:05] I tend to see trends

Speaker3: [01:00:07] In terms of how people build their architecture. So check this channel out and hopefully it will help. Thank you all so much. This has been extremely helpful. I have some good, good stuff to go through now.

Harpreet: [01:00:23] It's got a lot of good videos on the road. We'll be checking that that out as well. Great discussion, guys think that so much to learn about engineering. That's right. You guys are amazing.

Speaker7: [01:00:34] And actually, I'll go ahead and draw up a few more links in the chat for things that I think are really good because honestly, like the systems design portion of like when I was saying like the Ml Merrill Lynch email ops interviews was like almost the hardest. It was the most interesting, but it was also like the hardest part to prepare for because it's like you need to have a couple of layers of understanding, like there's tooling and sort of what tooling provides, right? So but just if you separate that layer off, you have to have an understanding of like irrespective of of whether you use a cloud vendor like what are the main components of what you need, like what's the minimum viable components of what you need? And then you still have to factor in like what are your what are your like resources and ability to like, implement, which I think is like. Like so there's understanding like what the cloud providers provide and usually they kind of make it

Harpreet: [01:01:29] A bit clear on their

Speaker7: [01:01:31] Website so you can look at the documentation, but it's the whole like cloud vendor independent knowledge. I thought that was really the hardest part in a way to prepare for because. A company that has like a very mature engineering culture and has like a bunch of like staff and up and principal staff, principals and whatever who have built the architecture from scratch and they're processing for millions of customers has just a way [01:02:00] different needs than like. A small tech startup where it's only like 10 to 15 engineers, half of whom are very junior right, and then it's like, you still have to. Oh, and even when you build it, you still have to think about maintenance. That's like a huge part of the conversations we have at work. It's like you can. It's always easy to find someone to build the project because that just looks good on your resume. Straight up, like career wise, building always looks good. You will always find someone who will build it. If you can say this is going to be like hot stuff, but the maintenance cost that is like the hard part to get teams to eat. Like so it's I thought it was always really, really complicated. Yeah, but I'll also post some links to to stuff that helps me at least like all the animal systems part.

Speaker5: [01:02:53] I got to say at some level, I totally agree with Akiko, there's this. There are these fundamental engineering systems design principles

Harpreet: [01:03:01] That

Speaker5: [01:03:02] It spreads the thinking in that principle. It spreads beyond the specific tools that you're using, right? And if you follow those principles, this like it's much stronger knowing those principles and really understanding that. And you can learn that with any system, right? Like for me, I think I went through this weird journey in the last 12 months where I'm probably stretching out into the late last year as well, where I started off down this MLPs and engineering side of things and trying to design at my previous company, I was designing this Data flow through to testing like the model training model, training automatically kind of thing. I've lost my word say, you know, just this whole, this whole pipeline of collecting the data, cleaning it, you know, training the model, testing the model and then putting it somewhere where the company can use it right and not knowing the tools wasn't really a barrier. It was more understanding

Harpreet: [01:03:59] What are the things [01:04:00] that'll

Speaker5: [01:04:00] Work and things that won't work out of event driven systems work differently to, you know, just connected systems at work. How do you how do you have services versus policy far off or manually? How maintainable is it? And I built the entire system of like Python and Docker and a very simple skills like not even a SQL server, just like SQL database sitting on local disk, right? And I understand how flawed that is and how broken that is.

Harpreet: [01:04:29] But one of my

Speaker5: [01:04:29] Big wins was actually talking to the guys that I left the company since. But I left in the hands of one of the graduate engineers who wanted to skill up a bit more in the ML side. And one of the big wins was that he was able to bring home the project wins two three months after I'd left without much overhead of like maintenance, right? And though that might not, that might mean that, yes, I haven't I hadn't used like, you know, top of the line tools and the correct tools for the job necessarily. But it reminded me that, okay, your engineering principles are still strong, you know what I mean? Like, obviously, there's a lot that I've got

Harpreet: [01:05:06] To learn, but

Speaker5: [01:05:07] Those basics, foundational basics were we're followed and adhered to made it quite easy for them to maintain and continue using. So yeah, you're absolutely right. It's the principles, not even the tools. And I found those principles were really. It was really easy for me to pick up GCP in the last three months because the same principles, it's just in a slightly different tooling, you know?

Harpreet: [01:05:30] Yes, I guess it's the case, we're just kind of understanding the problem to be solved right and then working backwards to whatever tool might be might be useful. Great discussion, man, you guys. Give me engineer envy here, I wish I was more competent on that front. I have zero.

Speaker6: [01:05:53] Oh, no, you don't. I'm just joking.

Harpreet: [01:05:56] Yeah, middle school, yeah, look at the post [01:06:00] up ahead and queue for question. If you still have a question, go for it.

Speaker5: [01:06:07] Yeah, just a quick thing, actually before before I ask that I wanted to thank Joe, actually. A few months ago, you like you talked about mental models on this, on this same podcast, and then I kind of went down the rabbit hole of reading up on him and listening about it and stuff. And one of the ones that I picked up that kind of addresses Mark you were talking

Harpreet: [01:06:25] About like focus as opposed to, you know, following

Speaker5: [01:06:28] Shiny objects,

Harpreet: [01:06:29] Right?

Speaker5: [01:06:30] In the last couple of months, I've started using a Eisenhauer matrix. If you've heard of that and factoring in urgency along with importance, like being able to look at that as a two dimensional Data, I literally like I draw one every single morning to make sure that I'm like on song with stuff, right? And this is like an evolution on top of the like the pocket book idea that if you listen to Kevin O'Leary, right, Mr. Wonderful.

Harpreet: [01:06:58] He keeps his

Speaker5: [01:06:58] Pocket book of just like the to do list, and he only does the top thing on the list. Everything else is like, I don't care, I will go to sleep and do my fitness and everything else, right? I've kind of taken that and I'm like, OK, how do I make that fit this two dimensional view as opposed to a one dimensional list to you? And I'm finding that in the last three or four weeks

Harpreet: [01:07:16] Has been this

Speaker5: [01:07:17] Return to a bit more disciplined, focused, right? So I'm trying to ride that wave and try to just keep tweaking these little things. So take a look at that. If you haven't seen it and have actually matrices before, it's a really cool concept. So, yeah, thanks, Jeff.

Speaker6: [01:07:32] That's so awesome. Yeah, good. Glad you found something useful in there. That's cool. Yeah, it's weird. The productivity stuff is there's no shortage. I think it's just what works with you too, right? So there's no shortage of books or ideas about it. But at the end of the day, I don't think it's the one size fits all. But, you know, Eisenhower apparently works for you, and that's awesome. So you keep doing it. Kevin O'Leary just needs this. I don't know why he bothered. This is one thing on it, but that's cool.

Speaker5: [01:07:57] So I think he also suffers from [01:08:00] the same thing, and that's where he's got like a twenty cool ideas that he wants to one day work on.

Harpreet: [01:08:04] But you know,

Speaker5: [01:08:06] It's an interesting and interesting comment on focus. But my question actually is more relevant to what Kerr was

Harpreet: [01:08:12] Talking about regarding

Speaker5: [01:08:14] There being a lot of side entries into the Data science field, right? I'm finding myself heading into a more of a mentorship leadership, people, some kind of space, right? And so far, like whatever, I'm mentoring younger engineers

Harpreet: [01:08:28] That I say younger.

Speaker5: [01:08:30] I'm still like a baby compared to a lot of people here. But a lot of younger engineers, when they come with the strong software engineering background, it's really quite easy for me to coach them on. Okay? These are these principles. So you haven't really thought of, let's, you know, revise some of those things as we design

Harpreet: [01:08:46] A structure, but

Speaker5: [01:08:49] That I feel a little bit out of depth when I'm dealing with people that don't necessarily come with that full engineering background because maybe like you guys have any advice on how do you coach people that don't come from that technical strength? So a lot of people go through door muscles and Data start the fantastical Data science, but there are still elements missing from their foundational stuff, like their confidence, just their confidence with get their confidence with breaking a system, right? And it's not everyone. It's like some cases. How do you manage mentoring people like that? Any tips, any advice? I'm just looking to grow?

Harpreet: [01:09:25] Market in Mexico.

Speaker3: [01:09:28] Yeah, I think Mickey should probably go first because you called her out first. Oh, I go, OK. All right.

Harpreet: [01:09:35] Well, I was going

Speaker3: [01:09:35] To say is this is from my experience. I did my internship a couple of months ago, and the similar thing wasn't for engineering, mainly for Data science, but a lot of the individuals are like completely new to coding and everything. And a huge mistake I made was. Caring too much that makes sense, and you're supposed to care, lie, you want to see him succeed, but you can't care more than them. And the challenge [01:10:00] is you have to be comfortable letting them fail because that's where a lot of learning is at. And so I want them to succeed so bad. So I'll go out of my way to make sure they wouldn't fail to learn these things and is putting it way more work than I should have. And looking back, I probably stole a lot of learning opportunities from them because I cared so much, right? And so just misplaced attention. And so I think

Harpreet: [01:10:23] One of the key things is like

Speaker3: [01:10:25] Figuring out where that gap is and for that learning to get to that next step and what will be the stretch thing to help them struggle through that while also knowing that you're there to support them.

Speaker7: [01:10:40] So we kind of have we kind of had this challenge as well at. Where I'm at right now, because. Part of the suit is I think it's just sort of it's a shift that you see when a company becomes more mature and at the same time that you see across the industry, right, which is that used to see increasing specialization in a lot of different roles and very specifically, so for example, like at MailChimp, it had a very strong engineering culture. So it was just sort of expected that everyone come in as an engineer. So the training skills was never it wasn't a question, right? Because they literally hired and actually the they had the engineering team first. Then they had the alliance and then the ML engine team was split out of Data scientists, and it was composed of Data scientists who happened to be very engineering. But. It's a challenge, right, because when you talk to like the science org, you know, and you ask them, what what is your KPI? What is your North Star like if you had to choose? If I gave you one hundred bucks and you could spend it across like [01:12:00] a basket of attributes? What would you spend it on? Right. And they're like, why would we spend on engineering when we have you guys there like we want to spend it on research because that's something that you engineers can't do. We want to spend it on statistical theory and research physical knowledge and expertize because once again, that's like that's something that yells wouldn't have that. We want to spend it on initiative on interest in cutting edge like deep learning algorithms.

Speaker7: [01:12:34] And they're like, why would we spend it on there? Like minimum viable engineering? That's what we want to spend it on. So what we did and this is I don't know if this is the best approach, but so far it seems like people are happier with it. Is we first kind of defined what were the minimum viable engineering skills that injuring skills that they actually needed? And it kind of comes down to like get so for being able to write Python code reasonably well, not like the level that if you know, if you were if you were producing like a flask like website or whatever web app, you don't need to do it that level. You just need to be able to write decent like feature engineering, model training, et cetera. So writing the decent code, being able to write tests, both unit and integration. And we've like we've done workshops on how to write tests. We record it. We put it up on a nice confidence page. Oh, what else? Being able to like work with with the tools that we have in-house reasonably well and work with other teams. And I think those are like and I mentioned get I think those were like the three main things that we were basically like because almost everything else, like for the cloud, for like the the cloud environment, for deploying it off as a service like we have teams that specifically work on that. And we just try to figure out how to make it easy on them.

Speaker7: [01:13:58] So I would say, like first define [01:14:00] like what that minimum viable engineering skill set that they actually need. And then I think the second part that we also did was we put together like a learning path that is not teaching science how to be engineers, but teaching the scientists the necessary engineering skills. And we basically curated a very, very select set of resources that we were like. This matches the values and quality we want to. You know, we want in the company, and then we just made it very clear, like the other thing we also did to is each day a scientist has to do like a cookie cutter project, so they have their first project that they need to go through. It's actually better to get them on that sooner rather than later. It doesn't have to be a big one. It can be a little small, even like a dummy regression model. And we just walk them through how to ship it because I think that is like. I think there's a couple people to kind of freak out because there's so much information so they don't know what to focus on. And then it's like if you focus them on what they need and you integrate and start day to day. It's a lot more. I think it's a lot more effective because reality, right? So you're not hiring data scientists to be engineers. At the end of day, you're hiring them to be data scientists, right? Unless you're specifically hiring more engineers, then you're kind of hiring data scientist who can code really at an entry level.

Speaker5: [01:15:20] So I totally, totally agree. And I think that the just keeping front of mind that that minimum viable skill set is kind of important.

Harpreet: [01:15:29] I think in my mind

Speaker5: [01:15:31] That I have some loose idea of what that might be given the kind of work that we're doing at the moment. But it's

Harpreet: [01:15:39] Really is it clear

Speaker5: [01:15:41] On what that learning path needs to look like and absolutely like what Mark said. I think part of it is just me, like wanting them to fail safely, right?

Harpreet: [01:15:51] Like, I just got

Speaker5: [01:15:51] To let them fail. Let them let them fail. Like, this is almost counterintuitive to what I said before, where software is breakable and easily fixable. Right? We [01:16:00] can go get checkouts. All right. So, yeah, absolutely. I think I'm going to take both of those things away and try to define a bit more of a mess for this. But I guess in almost a follow up is like, so we're kind

Harpreet: [01:16:16] Of we've got too

Speaker5: [01:16:17] Much operating, right? One is there's a product side where we're looking at a more engineering focus than the data science exploration side. And then we've got a few of our engineers that are more on the,

Harpreet: [01:16:29] You know, project

Speaker5: [01:16:31] Side. So the what I'm talking about the the professional services side, right? They'll be running smaller, smaller projects. Some of them will be like Mythbusters, and they're more data science and science. Some of them will be more, you know, small trials and small, small scale pilots and things like that, right? So there's a weird balance of Data science knowledge, but also a need for some element

Harpreet: [01:16:53] Of engineering

Speaker5: [01:16:54] Skills in-house,

Harpreet: [01:16:56] Right?

Speaker5: [01:16:57] And something you said about introducing them to a small project sooner rather than later. We're actually going the other way, we're training them up in the product team because there's a more stable team that can support our newer engineers and then moving them into the professional services side. But I mean, how does that kind of I'm not just trying to think out loud here almost is that kind of map's opposite to what you were saying.

Harpreet: [01:17:24] It's like a small,

Speaker5: [01:17:26] Isolated project that they can do is a cookie cutter dummy, maybe, and then grow into the rest of that.

Speaker7: [01:17:33] It's less about how small the project is, and it's more about the controlling, how much sensory input they're getting and the psychological safety aspect. So like, OK, so for example, I was at a company where I saw someone they had. Right, and there's a phrase, right, like if someone if someone screws something up and engineering right, it's not [01:18:00] it's not the individual, it's the process. Right. So something had happened. And. You know, screwed up the main like, get repo. Something broke in production. It was able to get rolled back. And they were let go afterwards. And. At that time, I think, and the reality is, so for something like it, I think most people learn it by poking. I don't think they actually tend to learn the underlying. They don't learn the structure, the abstraction, and they don't tend to learn the underlying principles of it. Because once once you understand the underlying concept of good, like, it's like, Oh yeah, you can just go and then you just have to like, look up the commands, right? But people would be so scared of deleting stuff from get in. And it's just that's just not really the way it works, right? You'd have to, in a way, try really hard. Once you commit and push something up, you'd have to really purposefully delete something and there's still tons of warnings and all that right? So but I know, like before I was able to learn the underlying concept that made me really scared to even like push code and contribute.

Speaker7: [01:19:11] And then once and then when I actually went through the missing semester course on Get, that was literally one of the best exclamations I have ever seen about, not necessarily like the implementation implementation details of get, but just how was it designed? How does it work or doesn't work? You know, and and I was like, Oh, this is this is great, and I felt less scared. So I'm like, OK, now I can. I can create PR and I can do code reviews and all that other stuff. But I think sometimes, like with more junior people, it's like they don't know what they don't know, right? Like. Like another way to look at this right is people who are first generation college students. And this was like came up as a relevant conversation because I guess in India [01:20:00] there's like a quota system where like there's a certain percentage of students who go to the IITs are from underprivileged backgrounds or groups within India. And so they get coding and then there's a certain quota that need be passed every single year. It's like scheduled

Speaker3: [01:20:17] Class schedule group, single class groups. Yeah.

Speaker7: [01:20:20] Right, right. And like, we have a similar issue. We have similar challenges with people who are first generation college students in the U.S. or their English language learners, right? Is that there's this idea that, OK, let's just get them in and then let's just give them free tutoring. And if they don't make it, it's sink or swim right, then they don't deserve to kind of progress. And I think the reality is that there's a lot of this like cultural capital and a lot of professional capital that people don't realize

Harpreet: [01:20:47] You need to

Speaker7: [01:20:47] Help people develop. Like I'm sure the raw motivation is there, but it's creating an environment of like that psychologically safe where people know kind of what the next steps are. And then you can be guided in a very intentional way to building up that skills they like, they can still fail, obviously, and you're going to run into challenges. But you know, it's kind of like at some point, if you have so many people failing or struggling, it's not the individual right. It's it's how you're supporting them, are not supporting them, right?

Speaker5: [01:21:23] Totally, totally. And I completely resonate with that idea of like if someone, particularly someone knew, if anyone really someone fails, that's really the process, right? I mean, I've messed up in my first job and that was on a manufacturing line, right? And we we fix the process. We found that there was a flaw in the protocol and setting up the manufacturing line and we went and set it up. And it was because it was because of how we conducted experiments on the manufacturing lines itself. So if we were putting a new robot onto it, there was, you know, there was a protocol that was missing or step in the protocol that was missing, right? So there really was. [01:22:00] It really is true that like, we've got to make people comfortable to fail. And I kind of agree is that with the the whole idea of really understanding the first principles of yet. I was lucky enough in my previous job to work

Harpreet: [01:22:14] With a lead

Speaker5: [01:22:16] Engineer who's got vast, vast amount of experience like he comes from

Harpreet: [01:22:19] Pre pre

Speaker5: [01:22:21] Days, right? This guy's real old school, right? Like I professionally old school. And he very, very simply explained to me the piece of paper and a pen. The plain concepts of just what it is and what it does. And then after that, learning how to rebase and stuff like that, like previous out my after rebase. Oh, no. Right? Like, like I want a linear, linear thing, I'll squash my comments and merge it in, but I don't want to reveal some edge conflicts used to scare me. And now I'm just like cherry picking. Absolutely. Oh my god. Cherry picking is painful, but it's not like now. I'm like, You know, it's not the end of the world, right? So now I'm really trying to figure out and maybe the missing semester, maybe that's the hot tip. And this is point them in the direction for that. I've just been

Harpreet: [01:23:07] Giving them kind of like articles

Speaker5: [01:23:09] On, Oh, here's this get here's how get rebasing works, and here's how this other thing works. But maybe a more structured thing would help them understand, because if you understand, you understand that anything you do is fixable basically in code.

Speaker7: [01:23:22] It really is a meta skill because it's it doesn't matter what kind of engineering you are at some point you're going to commit code. It really is like the it is the de facto communication standard between engineers of any culture. Race, religion around the world.

Speaker5: [01:23:40] You might call it you might call it something different, like if you're developing mechanical design stuff, the software version control stuff. It's mission control. It's not yet, but there is a particular principle behind how it works. And if you understand how that works, it's it's fun. Yeah, kind of. To close this off, it's kind of strange to me how little [01:24:00] having like. I mean, it's funny because having gone through engineering, school and computer science, it's strange to me that they just kind of touch on, Oh yeah, use it for your projects if you want. And, you know, unit test, maybe a little bit, they'll walk you through some basic fundamentals, right? It's just strange to me that there's not this like what I went to to the school for robotics. We're doing it using material, right? And not every team used it for the project. All right. Maybe two teams, you just material. Nobody else did.

Speaker7: [01:24:33] Yeah, yeah, like one hundred percent. What I would do is I would identify like the minimum viable and skill set. Figure out what are the biggest gaps suggests one or two good resources for those like create a little like confidence page or whatever called like. The learning could be a learning path or something. You know, you don't want to toss too many resources at them and then kind of get them start on, on, on code. It doesn't even have to be like an isolated cookie cutter project. It could be like a feature or whatever. But, you know, just make the expectation really clear and also make it very clear, like how they can ask you for help and all that. So it's not like your hand holding them, but you're just like, you're setting up the right sort of pathway in the steps ahead of them. Yeah.

Speaker5: [01:25:20] So and I think the other side of it is something else that you mentioned was that I think we just kind of assumed a level of engineering knowledge and background, right, which we really shouldn't have assumed in like in hindsight with some people. It's just that assumption meant that we've now fallen to this trap where we haven't like we need to be breaking off tasks into much smaller chunks to that this it's easier for them to handle right, whereas a larger engineering task that would require some design thinking, you know, like it's about building a lot to that stage. So yeah, no, I really appreciate the advice there. I [01:26:00] want to take a lot of this away, and if anyone has any more advice on this, just be me any resources, please,

Harpreet: [01:26:05] Please, please.

Harpreet: [01:26:07] And if you guys want to hear more from Mexico, you can check her out on the Kenji podcast. Ken's nearest neighbors check out Monkey Go on the podcast, but if you did not know, she's also on my podcast over a year ago. So definitely check out that podcast, Mikey. Go so go ahead and tune into that. Just a shout out to my good friend. Andrew DeCurtis is in the chat on LinkedIn, which grad school together spent in many, many, many hours studying for actuarial exams together and failing them and getting back up and studying and passing them again. Andrew says as a teacher, he cares. Too much can be overwhelming for junior staff and says functional programing is a must.

Harpreet: [01:26:48] So Andrew, thank

Harpreet: [01:26:48] You so much. By the way, men haven't seen you in a few years. Thanks for tuning in, man. So we'll go ahead and wrap it up. Thanks, everybody for hanging out. That was the missing semester that Mickey was talking about. Be typing the missing semester. There's a whole series of lectures, one on version control, in particular. There's a link to that in the show notes there should be in the show notes when this is released and also on the LinkedIn post here. You know, I'll probably spend an hour tomorrow morning watching it. I mean. I'm realizing, I don't know, get as well as I should. So that's it for this one. Guys will begin to wrap it up. Be sure to

Harpreet: [01:27:27] Tune in next week.

Harpreet: [01:27:30] I've got a few live events, right? I'll be live with Comet Imelda. Yeah, the session we're doing with folks from Uber, the real, real and work fusion.

Harpreet: [01:27:42] I'll also be

Harpreet: [01:27:43] Going live with Nick Singh and the gentleman from one salting. So there'll be a lot of fun. Also shout out if anybody is available on December 17th to take over the reins for

Harpreet: [01:27:56] Me

Harpreet: [01:27:58] For four happy hours. Let [01:28:00] me know if not, no worries. We'll just have to keep it a short, happy hour on December 17th. Maybe, you know, keep it capped like forty five minutes or so because I'm going to a hockey game first one in forever. Thanks, guys, for hanging out with me. Thanks for, you know, just being here. I really appreciate all of you guys and just all of your support over the last year of this podcast. As we crossed that hundred thousand download mark could not have done it with all of you without all of you guys, take care of my friends. Remember you got one life on this planet? Why not try do some big cheers, everyone?