HH-09-Jul-21.mp3

Harpreet: [00:00:03] I what's up, everybody? Welcome. Welcome to the @TheArtistsOfDataScience Happy Hour at number 40. This is the fortieth week in a row we've been doing this man. I can't believe that 40 weeks in a row we've been here live and direct man, answer you guys his questions. Building this community couldn't be happier to be here with you guys every single Friday. Shout out to everybody trickling into the room. We have a man can in the house can. What's up? We got Antonio. Antonio, man good to see you again, Russell, my friend. How's it going, man? We got we have investors. They're entering into the room. Nesha, what's going on now? Superexcited to have all of you guys here. Hope you guys had a good week. Can't believe it's already damp. First of all, I can't believe I do this for 40 weeks. It's insane to me. It's just this wild another another few months and it'll be a year straight. So that's that's cool, man. Um, well, hopefully, as the week has been good, got a chance to hang out with my friend Mike Delgado on the Data Talk podcast earlier this week. That was a really special episode for me. Mike was the first person that I had reached out to, one of the first people that did reach out to to be on my podcast.

Harpreet: [00:01:13] And I actually went through his list of people who he had on his podcast. And I was like, hey, if this person's been on one podcast, maybe they might want to be on another one. And so that's how I kind of poached his audience list, or at least an episode today with Barbara Oakley. She wrote the book, A Mind of her numbers. She is the teacher behind the course, um, learning how to learn the most popular music on Coursera. So that's a great episode. Really, really sad today. But we recorded it on January 3rd, so it's quite a bit delayed. But, yeah, I'm super excited. How do you guys hear if you guys have questions, wherever you are, whether it is on YouTube, Twitch or LinkedIn, go ahead and leave your question [00:02:00] right there in the chat and we will get to it. There's also, um, there's also a link wherever you are to get into the room as well. So do join in on that. What's going on, man? Super happy to, uh, to to have you here as well. And is asking what's been my favorite podcast episode. I've done it, Antônio, in terms of like ones I've been on or ones that I've recorded

Antônio: [00:02:25] That you recorded.

Harpreet: [00:02:27] Yeah, I mean, two of my top favorite ones that I've recorded, without a doubt, it's got to be within just that and kanji. But barring those to answer, barring those two, I'd say, um, uh, do you mind everyone's having been released yet, but the one of the ones that been released, definitely Robert Green, that was a really cool episode to do. Next week I got one release with James Altucher, and that's one of my favorite ones as well as I really enjoyed. Enjoy that. Matt Damon, what's going on and do. Yeah, I forgot to have interviewed Annie Duke as well. That's crazy about these interesting people ever had a chance to interview. But yeah, let's go ahead and get this thing kicked off. You guys got questions. Let me know right there in the chat. Wherever you are, I'll be monitoring all platforms. But let's get started with the with the with a warm up question. Right. So how about this? So we've, uh, man, I should have thought about this before I went live. How about this? Talk to me about, um, for those of you who know how to do NLP or have been learning about NLP, um, would you guys be able to share some resources or share some tips with that? With us? And by us, I mean, uh, mostly because I'm trying to learn that as well. Uh, everybody here has an experience with that. Let me know. Um, in the meantime, I'll be looking for questions in the chat and on LinkedIn and everywhere else. Um, but if you don't have any any insight into that, then go for it and ask a question. Rasselas he had a question here, the chat. Go for it or [00:04:00] else.

Antônio: [00:04:03] So yeah, I'm just interested to know from 40 episodes. And so that's a lot. What's been the most interesting mess up in all of that time or what have you learned the most from

Harpreet: [00:04:12] The most interesting mess up in terms of. I just I mean,

Antônio: [00:04:17] We had just something went wrong.

Harpreet: [00:04:18] Oh, man. I mean, I feel like the beginning parts of most episodes go kind of wobbly because I'm trying to get the conversation going and started. Um, I mean, I remember there's one week where I was talking and I think I left my microphone on mute for like thirty seconds or something like that. And that was right around April ish or something like that. But I was just looking through some of the back catalog of the past forty episodes. And there's there's a period of time and there's like forty to fifty people in these things right around winter time. And that was a lot of fun man. Um, some of the early ones like No. Ten, eleven, nine, ten, eleven have a lot of great inside information. Um, those are just ones I was looking at in particular. For for today, be a shout out to everybody on LinkedIn YouTube twitch monitoring the comments. So if you guys have any questions, please do let me know. Yeah, the netlist. Let's turn to the let's turn to you guys. Man How you guys been doing? Ben, what's going on, man? Nice shirt, by the way. I like that.

Speaker3: [00:05:18] Yeah, we coordinated rate. We called before. It's a little shirt off, right.

Harpreet: [00:05:23] I love it and I love it.

Speaker4: [00:05:24] I usually practice, but I'm off my game today then.

Harpreet: [00:05:28] But practice just just be an awesome practice.

Speaker4: [00:05:33] A little hot Friday.

Harpreet: [00:05:34] A little Friday. Is that necessary?

Speaker4: [00:05:36] Yeah, every every job I've ever been at, I instigated Aloha from our institute and I don't know exactly what the right word is there, but a good way to pick up morale. You know who looks unhappy in a Hawaiian shirt over?

Harpreet: [00:05:50] Yeah. For a man I didn't know Aloha Friday, it was a thing. It's just I just coincidentally happen that I'm always wearing these.

Speaker3: [00:05:57] It is. Yes. We have a song about it back home. There [00:06:00] is an Aloha Friday song.

Harpreet: [00:06:02] So then are you from from Hawaii

Speaker3: [00:06:04] And

Harpreet: [00:06:05] No shedid when you when you move to the mainland.

Speaker3: [00:06:08] Yeah, sure. It's official. This is this has been approved. Hawaiian shirt.

Harpreet: [00:06:13] Dude, that's awesome. Yeah. I did not know that you were actually from Hawaii is pretty cool man. Um Bayamon. So let's go ahead man and then and see if there's any questions in the chat. Um so Russell asking biggest mess ups, how about this one not having a a question lined up for four at the beginning of happy hour. I've been just I've been grinding all day and I've been really trying to learn natural language processing, mostly because I'm sitting on just a huge amount of text Data from the transcripts of all the previous podcast, all the previous happy hours, comment about office hours, all that stuff. I feel like there could be something that that I could do with all that Data. Um, so I started cleaning a lot of it. Um, but it's really, really tedious. So I'm trying to outsource that, looking for people on up work to clean up those transcripts because the automatic translation gets about 90 percent of it. Can I say you say yeah.

Speaker4: [00:07:13] Well two things then. I didn't know you're from from out here. I live in Oahu now. So if you ever back out this way, let me know. The second thing I know, one of my our friends, Abhishek Thakkar, is working at hugging Bass. So they're doing auto NLP could be a cool product to look into the platform. He is an absolute rock star, definitely knows what he's doing on that front. So he also might be a good person to reach out to to chat with the really good guy and always looking to help.

Harpreet: [00:07:44] Yeah, he's the Kaggle grandmaster guy, right?

Speaker4: [00:07:47] The first four time grandmaster.

Harpreet: [00:07:49] Yeah. Yeah, he's he's got he's got a lot of awesome stuff. I didn't know he was out there hugging face working on a product like that. I've got to go check that out. Um, Batman this. Let's turn it over to the [00:08:00] audience. Guys, help me out here. If you guys have any questions, please do let me know. It has been a long, long week for me. So, uh, my brain is a bit fried.

Antônio: [00:08:08] I have a question for you guys. I guess for everybody that's creating content. What was that moment or can I not win that where you kind of got over the hump and so, like, right the day I actually started recording this podcast or because I know for me I started recording some stuff, but it was quite a few months where I'm like, well, I'll I'll start and I'll do it tomorrow. But then something inside you, you know, doesn't. And then one day it just just clicks. So, um, but I'm still not there. Yeah. Like you've been doing this for 40 weeks straight. I guess it's a habit for you now, but if you could share a little bit about kind of like what is that that got you to take action.

Harpreet: [00:08:53] Yeah. Yeah I do. That was that way. When I first started the podcast, I was just I was thrashing a lot. I got recorded a bunch of episodes, but then I was like, I don't know what the format is going to be like. I wanted to do it like like a talk show. It would be like part opening monologue. Then it would be the interview then would be like like audience questions, stuff like that. I had like this thing planned out and, um, I just noticed that I was just wasting time writing monologues. Right. And and just I was like, do this is not useful. It's not getting anything done. I felt like twelve episodes recorded. I need to push these out and could do something with the otherwise the momentum was going to die. Um, so it was just then I was like, yeah well I know for a fact what I'm doing. This is not conducive to my ultimate goal, which is just to put stuff out there. Right. Um, and I just said, fuck it, I'm just in to do interview shows like that's that's what I'm doing and I'm going for. And at that moment I just kind of let it rip. And then from there, just trajectories. I mean, insane can read about you man, because you're like almost said we're talking yesterday for a while and it was cool. You [00:10:00] did like a hundred and thirty videos, radio, two hundred videos, something like that.

Speaker4: [00:10:04] I think I'm almost at two hundred. And when we actually talked a little bit about this yesterday, I think it's pretty important is that once you start building momentum, that's a beautiful thing and you can continue from that. But in order to create momentum, you need volume. And I think you should produce good stuff out there. Like you don't want to put out stuff you don't to just, like, throw a bunch of stuff out there and it all be trash. But if you can get to a point where you're producing consistently, where you've got a good backlog, a lot of content, that's where you grow the most. One of my favorite things, it was either like an art class or like a photography class. Right. And the professor essentially said that half of the class is tasked with taking one picture and making it as good as possible. That's what they'll be graded on. And the other half of the class is going to be evaluated on how many pictures that they take just purely on volume. And it turned out that the half of the class in both of these scenarios that were in charge of volume, just putting stuff out there, they took significantly better pictures and the pottery class, they made significantly better pottery because they were getting the reps in. So if we're looking at content creation as this consistent journey, this iterative process. That is how you create that momentum. That's how you get to that point, like now it's just like I look at every piece of content that I create as a way that I'm improving my future content.

Speaker4: [00:11:26] So it isn't this like, oh, I have to get to this threshold every every single know, whether it's a newsletter or whether it's my podcast, whether it's a YouTube video, it's all just like part of this moving train that's going forward. And I think that we got a little bit too wrapped up in the early stages, too, because nobody nobody looks at our stuff in the early stages. So you can like, frankly, make the quality not quite as good, but make sure it is practiced, make sure you're getting those reps in, because when people when it is starting to get traction, when you are getting viewership, [00:12:00] you'll be glad that you put in those hours before that. No one is watching your stuff when you are making the mistakes. And that's OK. Right? Like if someone sees that I made a YouTube video two years ago and it frankly wasn't that good. The editing is crap. Hopefully the advice and stuff is good. They'll see a video that I've made now and be like, wow, this person has improved so much. That's what they're going to be evaluating. Guess it's not going to be like, oh, we made this video two and a half years ago and it was trash. They'll be like, wow, look at his new video versus this one. This person has come so far, so far along the way and they get to be a part of that journey, which I think is is incredible as well.

Antônio: [00:12:35] I like that. Thanks for sharing.

Harpreet: [00:12:37] Absolutely. Do I love that, too? I mean, just thinking about, like, the process of creating, do you find yourself like like thrashing like purposely doing stuff that isn't moving the needle forward just because you kind of feel stuck? Like for me, for the longest time it was editing, like I used editing my podcast as a distraction for other things related to my podcast. And then it I just finally realized that, OK, I can't waste my time doing this because this particular task isn't going to really impact the output of the quality of the podcast, which for me was because I need to ask good questions. I need to research the gas, I need to understand their body of knowledge, like that's where I should spend my time, not editing the podcast and stuff like that. You just kind of tend to outsource. Did you come across any issues like that? Can or Vinn or anybody else like that creates content.

Speaker4: [00:13:27] Eleven go first and then I definitely have some thoughts on that one.

Harpreet: [00:13:29] Yeah, definitely

Speaker3: [00:13:30] No. You carry the momentum. I'll show it afterwards.

Speaker4: [00:13:33] Sounds good. So one thing that I've I've found is I also have that problem right there. Plenty of things that I don't see as much value in. And I know that I probably should do that or have to do them or whatever that might be. About a year ago, I think it was twenty nineteen, I decided to devote all of my attention to YouTube. Essentially, that would be the platform that I focused everything on and everything else would essentially just be [00:14:00] right if I gained LinkedIn followers, whatever that might be, if I created an audience there, it wouldn't be because of the dedicated effort. I put it on LinkedIn it would be an after effect and a positive externality from the stuff I was doing on YouTube. And so essentially I was able to use that tunnel vision to focus on what exactly I needed to do on YouTube to grow to me, being able to narrow it down, use like an 80 20 principle or one thing principle where you're like, hey, these are the things that are going to move the lever. The furthest I need to focus on. Like, if I'm if I were on a boat and one motor is pushing a lot harder and it moves the boat a lot further, we're going to focus our effort on that motor rather than the other ones that aren't as relevant that are going to move the ship quite as far.

Speaker4: [00:14:44] So I think that it's OK to eliminate it's OK to fall behind. Or if you have if you're doing a newsletter and you're doing a podcast and you're doing YouTube, like you should get a head start with one of those and it'll carry everything else. To me, it's it's not efficient or it's a bad idea to start all three of those at the same time and expect to have success in any of them. So I would always look to what you can trim down, what is like what's not going to detract from your brand, what's not going to detract from whatever you're putting out there. But also think about, hey, I'm going to focus my attention on what does move me to for this what does what is going to create the most impact. And you are inevitably going to have to sacrifice something, but you can make efficient sacrifices. They create momentum or that create huge surplus in other areas can.

Harpreet: [00:15:39] Thank you so much, man. That's if there's one person take advice from on that side then. Right. That can. Thank you so much. Then let's go to this here about this ad. The first question that Antonio was asking was all about getting over that hump of deciding to push the content out. What was that like for you?

Speaker3: [00:15:54] It was interesting. I started originally on Twitter in twenty. I think I started my account [00:16:00] in twenty ten and I actually started using my account intelligently in twenty twelve. Same thing with LinkedIn. I started actually using LinkedIn in twenty twelve, but I had that forever and the first thing I did was curation. I didn't think about it so much intentionally as to what I was doing with curation for me was a great way to start creating a voice for myself because I was using my ability, my filter and my ability to say, OK, this is important, this is excellent content. Here, check this out. This is why this is important. Was like my next thing then I was saying, OK, read this because it's important and here's why. And it was cyclical. I started developing my voice through going sort of standing behind other people first and then explaining pieces of why this is important second. And then that pushed me to have enough confidence to start posting on a regular basis on Twitter and then on LinkedIn. And so that was the gateway for me, was it went from curation to find my own voice and then realizing that people actually wanted to hear my opinion. And that was the interesting thing, was going from why would anyone care about what I have to say to, oh, well, why do so many people care what I have to say? And it was that process of discovery for me that I realized people wanted to listen to me and then I had to go backwards and figure out why they wanted to listen to me.

Speaker3: [00:17:27] And then I work backwards to who was listening to me, what was I actually influencing? Because I ran into somebody who probably didn't want me to drop his name, but I ran into somebody who said, OK, define influencer. And that was one of those. Oh, well, I guess you would have to influence people to do something that you'd actually have to get people to act, get people to think, get people to do. And that started me on the journey of, OK, so I want to influence people. What do I want to influence them to do? What [00:18:00] do I want to teach them? What do I want to expose that? I don't think people are talking about enough what parts of the field are undercover, what parts of the field and what people in the field. And you can hear this evolution. And this was really over the course of seven years going from doing this unintentionally to doing this more intentionally, picking the topics that I thought were. Are worth going after, and now I'm at the point where I really have some fairly well defined lanes, I have the channels that I use just opened up YouTube because I think a lot of junior level and early career level Data scientists like YouTube. And I think I can also teach some more complex concepts using video than I could using blogs, which is where I started out.

Speaker3: [00:18:52] And I see things like LinkedIn sort of falling behind in reaching people and as a creator platform and hopefully they'll be fixing that. And that's another piece of the puzzle is looking at how all of these platforms have evolved and some have fallen behind and being ready to let go of one hundred and something thousand followers on LinkedIn. And I'm probably going to be moving away from LinkedIn unless they move their needle forward for getting people involved in content creation. And so now I'm looking at substory and having a different relationship with newsletter's. And so the final piece that will say is, if I could do one thing better, I would have figured out how I was going to monetize from the beginning rather than getting a few years into my content creator career. And then in like twenty, seventeen, twenty, eighteen going, wait a minute, you mean you can make money doing this? OK, so how am I going to make money. And it was sort of backwards where if you're going to be a creator, always expect to [00:20:00] get paid. Don't feel bad about expecting to get paid for sharing your ideas, your knowledge and your abilities and figure out how to monetize up front and then keep those two strategies of providing a great resource to people in the community. But at the same time, there's a return. I should be getting some cash out of this and figure out how you're going to do that.

Harpreet: [00:20:23] That's one thing I was talking with Ken about yesterday as well, was figuring out ways to to, you know, get some some support and funding monetization. First for some of this effort is a lot of effort and content. Create content. It's definitely not not easy.

Speaker4: [00:20:39] I would push back against that a little bit. Sorry, but I actually don't think you you should get into content creation for the finances because I think you'll hit a threshold where like like money. The monetization of this comes at the long way into the journey, regardless of how you approach, at least from my experience. And if your your your goal is to get to that monetization threshold, you're going to have to put in hours and hours of work, hundreds of hours of work to be able to get there. And if it's I'm not saying that like, then I know money is not the only driver for you, but if you're your sole focus, is that monetization, if you're, let's say, our data scientist, like your hourly rate, my hourly rate for content creation still is like less than twenty dollars per hour. Right, for the amount of time that I put in. So if you're going down that route for the monetization purpose, really difficult road and you better for me. I had to freaking love it. I had to like that the community building and like some of these other stuff before I flipped on that monetization switch. Otherwise I wouldn't have had you know, I haven't had longevity out of doing this for a couple of years, but I don't think I would have been around as long [00:22:00] as I've been because I would have been pretty discouraged pretty fast.

Harpreet: [00:22:05] Yeah, but definitely not one hundred percent of anything on that comment there. No, I

Antônio: [00:22:10] Was going to say I was following something very interesting about YouTube, like the big influences that push multiple many videos and things like that that that kind of, you know, create so many videos and at one time they just scale exponentially. So whatever it is, in order to avoid and do so without knowing and the successful ones, the successful creators on YouTube, they avoid being one hit wonders because of that pile of videos that they've built over the years. So to the point of in and in can, this is actually what drives the most value. So you get to a point where you make that one video that gives you a million plus views, but people get tired quickly and you might become a one hit wonder. But when you have a portfolio of videos of subjects that people can go back, that's what really helps you drive your follower base. So it's really important to not focus on just making that one thing that gives you a lot of views, but really increasing that list of portfolio subjects, given that they do enhance your brand and they do address the issues that you want to address.

Harpreet: [00:23:23] Yeah, and excellent points, Antonio. Hopefully you got some. Some good tips there and hopefully start pushing some stuff out and can get content out there.

Antônio: [00:23:31] I mean, you know, it's been good, I guess, for me was so I spoke to one of my old college classes and a lot of the kids started asking about resumé advice. So I kind of made a video and I was like, all right, I'm going to keep repeating myself. Let me just send it out there. And the video did great. I posted it on Reddit and ended up picking up what in my first video within the now or like 600 lakes went up to like ten thousand views on YouTube within [00:24:00] a day. So which I wasn't looking for that. So I was like, all right, so let's started doing some more stuff. And while I do think resume building is important, I keep telling them, like, OK, once you how do you interview? Once you get the job, how do you succeed? I think one of my other videos is about networking after your first job. And I just got stuck at that point where all the people that were following me now just care about getting that one job. And I'm like, I like helping you with the job, but you got to realize that it's a very, very long journey. And it seemed to be the community or the people I was building was they will come to me, get a resume. Otherwise they get the job and they disappear off the face of the earth and they weren't kind of following the other stuff. So that's why I kind of had that a little bit of a down time trying to really see where to focus, because I do like I said, I like helping them with the resume, but I think I want to go beyond that. So I think I just got to keep going and just find the right audience. But at the same time, like you guys are saying, I guess if people just want a job, then I can help them with that, then maybe I should just roll with that.

Speaker4: [00:25:14] I'm actually always fighting that. So a lot of my content is for people in the early stages of Data science. Right. And I'm seeing the exact same thing. Once they get a job, they just don't watch as much content for me because that's what their initial focus is. And I think that it's OK to to even touch down within that and to say, hey, this is what our focus is, but to also have little tidbits that push that journey further. So, you know, on my through my content, I have some tutorial stuff. I also have some advice on the job stuff. And inevitably, after someone gets a job and they're already familiar with you, once they have a question related to that domain, they'll probably come back and take a look at it. So it might not be viral content. You might not get as many views. But I look at those things as longer term content, things that people [00:26:00] or something that's evergreen or people are going to continue to refer back to. And I wouldn't look at it in terms of chasing the audience. I think about it of the life cycle of these individual people and where you can make touch points with them. So like it might it might be a year down the road. It might be six years, six months down the road, but they probably will come back if they do have a question. And it's awesome if you have content for them. Right. When they're ready for it. So I would start thinking about it like, hey, let's go on the journey with them. Where do I want to be in their life and where does it make sense and make content around that, but also make content you want to make where you think the advice is going to be the best because you're going to love making that content. It's going to be so much more fun. So that's kind of where I would I would leave out like that.

Antônio: [00:26:45] And I'll thank you. That's that's great. Otherwise, maybe we can build predictive models, predict their next problems. And we'll also hear

Speaker4: [00:26:54] You can you can go through all the comment Data on my videos if you want. There's a lot of recommendations there.

Antônio: [00:27:00] Once Harp pre marsters NLP will get him

Harpreet: [00:27:03] And they'll be fine. They'll be fine rather than a great conversation. Everybody loved it. Thanks for uh thanks for. Help me kick it up Antonio. Appreciate that. Appreciate the question. Kick it off Christian. You've got a question where the Christian goes. He's still here.

Speaker5: [00:27:16] Yes. The cut off the video for a second there.

Harpreet: [00:27:19] But you're not actually driving right now are you. Because try to be so

Speaker5: [00:27:23] Good about just picking up groceries. I'm here with my son. And he said, hey, that guy has a really small voice. Just tell him some more videos that both. Now, my question is just around opportunities and solutions and frameworks, especially curious to hear from the Data standpoint, but you guys have favorite frameworks or approaches you like to use to develop solutions and or identify high priority opportunity areas. You think about it in terms of use cases. That's on my mind because I'm in product management right now. And the better I get at developing that [00:28:00] kind of opportunity solution space, the more I see how it applies to pretty much everything. Most important, step one seems to be a good opportunity. Right. So how do you guys go about doing that?

Harpreet: [00:28:10] Awesome question, man. Let's go to you for this one. Shout out to Eric Yo Yo.

Speaker5: [00:28:15] So for sizing up like use case opportunities that I'm not totally sure about, that's something that I've been at that I'm. Talking about a lot as well, and in kind of on a daily or at least a couple of times a week basis as I'm trying to size up different different priorities, I think I think for me the biggest thing is just like knowing I always am going to my stakeholders and trying to understand better. I just assume and figure that they understand the customer better than I understand the customer. And so I'm just always trying to how do I get into their get into their shoes better? And then mostly I think of frameworks I've primarily been involved in like process improvement framework. So like Dimmick for like Six Sigma and things like that. But I don't know a whole lot about before that, like the breaking down that defined to finer and finer pieces. So I'm listening.

Harpreet: [00:29:08] Yeah. Let's let's go to then for this one that after then we'll go to Greg also shout out to everybody else in the room as she looks up to them, except for my man. Good to see you back, J. Matt. Vin, what's your take on Christian's question about frameworks for identifying opportunities, use cases, things like that?

Speaker3: [00:29:29] Actually, I just did a video on this awesome crossover. There's one. This is an ancient framework and it's the concept of portfolio management. Things old don't use it in strategy planning, but the thought process is still really valid when you're looking at, OK, I have tons and tons of different things that I could possibly be working on. And you look at all of your opportunities. I'm assuming you've identified a good number of opportunities, things that you could potentially be working on. And I would say from a perspective of, [00:30:00] OK, I got opportunities. Now, that's a portfolio, basically. And so you want to evaluate all of those ideas based on their potential returns. And that means you're probably talking to a ton of different people who are going to give you opinions on the returns, and then you're going to have to go out and validate which ones which one of those opinions are actually true, which ones, which problems are worth solving. And you're going to gather a bunch of opinions. You're going to have to review your portfolio and you're going to have to make a decision based on revenue, an impact to the business.

Speaker3: [00:30:35] You're also going to have to take sentiment into account, all of those opinions and all those interviews and try to figure out, OK, this is the highest returning with also the highest backing because you need both. You need money that's significant, but you also need people to care because you can solve problems and say, hey, this can give us a ton of cash. But if no one cares, I don't understand why this is such a hard barrier to get through at the very beginning of machine learning in an organization. But you have to get people to care even if you tell them, hey, there's a ton of cash. So take those two pieces into account and apply a portfolio management strategy and evaluate them with respect to how much will they grow the business? What is the potential long term to grow the business, how much room to run, how much market share is available or how many how much cost savings total short term, and especially some projects have long term cost savings that are even bigger. So think of it that way, Greg.

Harpreet: [00:31:35] Let's hear from you then after Greg can. And if anybody else wants to jump in on this, please let me know. I'll add you to the queue or if you have a question, entry to the queue that goes for everybody on LinkedIn Twitch and YouTube. Greg, go for it.

Antônio: [00:31:49] Yeah, I'm I'm such a promoter of standardization. If you are in the product space, there's nothing better than creating [00:32:00] a standard funnel for hearing the voice of the customer in capturing that feedback as a mechanism to drive the roadmap for your product. So what you need to spend time on is building that mechanism. Where you taking in their feedback? And if it's a team of business, folks have them create or you need to create a matrix that ranks the the ideas, the business ideas, recommendations or pain points that they're providing to you. And they need to provide justifications. Oftentimes we think, oh, I'm a data scientist, I need to check what the valuation is, what the what the the cost or the benefits will be, what what is the revenue and things like that, how the business folks do it in your intake mechanism. Put all of these requirements put in template of questions that they need to ask. They need to read them to make sure that the problems they're bringing to you are in fact solvable by your solutions. So so have this mechanism built in in the line to give them training on how to fill it up.

Antônio: [00:33:17] Then you what you will do is in the beginning, it might be hard. You might find people fight it, but once you have it. It will be so much easier for you to have business folks to come in and put their ideas, put their pain points in there that will be scored and that you can append an effort to it because now you already know what the impact is to the business because they've already provided it to you once you open the effort. Now you have effort and impact. So you can kind of provide some sort of what you say, but you're going to trade effort with impact to see what you [00:34:00] need to go after. For example, you may see that tackling three low effort, low impact pain points is better than focusing high effort to a high impact idea or feature. So you need to make sure you have that process where you're gathering all of these pain points under one mechanism that makes it easy for you to really think about how you need to transform your product or your services that you're providing to your stakeholders. So think about that standardization and you'll be good

Harpreet: [00:34:34] And such fire advice for everyone. This is awesome. Thank you so much, Greg. Shout out to everybody else that joined us, Joe in the building. Good to see you, Joe. How are you doing, Ronit? What's going on, Dave, my friend? How's it going, man? Spencer's in the building. What's up, Spencer? What's going on? You guys got any questions? Please do. Let me know

Antônio: [00:34:55] If I can add on to that,

Harpreet: [00:34:58] Please. Absolutely.

Antônio: [00:34:59] Yeah. So so for to use cases like everyone, everything is spot on the Greg Bean chair. I think also the one thing and I think we've talked about this before, but it's always also like you have multiple use cases. Right. And a lot of them are going to meet you, let's say, I don't know, like one hundred million dollar opportunities, but they might not be easy or there are long term projects. And if you identify that I in my experience, it's not great to just focus on that, because if you wait too long, like I'm going to deliver to you this in like three years, the business is going to get discouraged. So you always want to find some low hanging fruit as well. So maybe find couple opportunities that are not going to do that. Kind of like huge hundred million dollar impact. Maybe they'll do a million dollars or whatever it is showing those quick wins and be like, hey, we did this in like three weeks. Let's see what we can do. So kind of keep them going. You feed them little pieces of bread, you know, [00:36:00] like the the dinner hour finally arrives. And I think that is very important to keep in mind because otherwise, with the way these companies work, it's like people move around and things change. So you don't want to wait five years or three to five years to just show like one project that you're going to deliver.

Harpreet: [00:36:20] Absolutely, man, and so

Antônio: [00:36:21] Much and I got distracted by Mr. Benwood, his new hairstyle. I see

Harpreet: [00:36:28] There's some new on on their

Speaker4: [00:36:30] Real life cyberpunk character.

Antônio: [00:36:32] That's right. I'm just trying to just trying to entertain the kids.

Harpreet: [00:36:36] That looks awesome. And yo, Matt Damon, I saw you trying to meet there a few times, my friend. Good to see. Yeah.

Antônio: [00:36:42] No, sorry is. Yeah, I can see you guys just to Antonia's for Data is getting more and more mentions in the public domain, obviously beyond the buzz words. But companies are mentioning in their quarterly earnings calls that Data science has some major effects on their earnings and human capital companies are talking about it. So the quick wins that until you just talked about, that's starting to build more and more. So I wouldn't be surprised if your counterparty starts to understand that there's actually some potential impacts for Data sites, that it's more than a conceptual thing these days. That's all I wanted to say.

Harpreet: [00:37:18] Thank you very much, Matt. Appreciate that. Man Speaking of Data, sounds like people keep talking about science is dead and it's dying or whatever. Ben, what are your thoughts on that?

Antônio: [00:37:30] Data Sciences Data that's not delivering value. Sayat a I'll get taxed where people say, hey, I'm thinking about firing my Data science team. But this is because it's been like many, many months with no value. And so I think that's the distinction I make. I think good data scientist understand the business side, the partner, which means they're not OK with 10, 12, 15 months of academic work. They have a sense of urgency. Yeah, it's definitely not that. If anything, I'd say it's expanding like name a department that doesn't need data science. I think that's really [00:38:00] hard. Name a job that doesn't need data science. It's like a job that doesn't need a database like Syckel database.

Harpreet: [00:38:07] So can let's hear from you and then we can do to you.

Speaker4: [00:38:11] Yeah. Honestly, this is a question I get asked a lot on the Internet. Is Data science saturated? Is automated going to take over what I'm doing? Thanks, Ben.

Harpreet: [00:38:21] But but,

Speaker4: [00:38:23] You know, to me, I think that this is a really silly thing. I think people get so excited about what they see, the companies who are just elite tech companies who are on the cutting edge, what they're doing. And they forget that like almost every other company. Ninety five percent of companies are lagging behind in their analytic capabilities. And a lot of the times you really do have to walk before you can run. You have to create good data infrastructure. You have to do a lot of these other things as prerequisites to performing data science. And so there's going to be this this lag of companies that are catching up to the baseline that are going to be needing to hire data engineers, data scientists, machine learning engineers, whatever it might be. And that's going to be over the course of the next ten, fifteen years. So companies, big companies, if anyone's worked in one, they move really frickin slow. And the idea that we're losing data science jobs aren't going to be any around, they're going to be taken over to me is ludicrous because, yes, technology moves fast, but people inherently don't move very fast, especially if they're associated with a very large organization. There's nothing to say that some of these organizations that are on the cutting edge aren't moving fast. They find unique ways or whatever works for them to be able to do that. But to me, there's a really hopeful message is that there's a lot of opportunity in Data science. And knowing that, hey, I might I might not work at a Google or an Amazon or whatever it might be, the Data science and those domains might be very different. But in an industry and a lot of these other places where they aren't initially focused on technology, I believe there's going to be an unbelievable amount of opportunity within this field or with what you can do [00:40:00] going forward. So I'll probably make a video on that at some point. But to me, it frustrates me to see that question because what we're seeing in a small, small segment of the industry is not representative of all the business.

Harpreet: [00:40:16] Yeah, I absolutely agree. And just the fact about big companies moving slow. Yes, they move slow. Very slow. Joe, let's hear from you.

Speaker6: [00:40:27] Yeah, I mean, Ben Taylor, remember back in the day when, like Data, scientists all thought that, like, analysts were going to be dead and it's a dying profession and all this stuff, because I think better name came up in the Data Machine Learning Epoch about the same time and in the same city, actually.

Antônio: [00:40:47] So Joe has aged a lot better than I have. His models were both eighteen years old. So it's crazy

Speaker6: [00:40:57] Stuff. But I mean, analysts were written off for dead in sequels, written off for dead. And I think that's actually the best time to get into something because that means the hype cycle is starting to wear off and now you're going to start producing more value. So whenever something is there enough or Data, I always get really skeptical and I think it's actually the best time to get into something, honestly, because that means all the posers and. People are kind of leading, so

Harpreet: [00:41:24] Do you think that in the next coming years that the skill sets that many Data scientist possess, for example, being able to code quantitative skills, reasoning, things like that, are going to be requirements for this new phase of white collar jobs that are emerging or no call type of jobs that are coming out that make sense.

Speaker6: [00:41:47] Kind of

Harpreet: [00:41:49] I guess what I'm trying to ask is, is everybody going to need to have some basic level of Data science skills going forward that now that Data [00:42:00] status everywhere, Data and everything, better to go for

Antônio: [00:42:03] It? They have to be able to scope a problem. They don't have to be able to solve it. So if I am like someone, if I'm working in marketing or in sales or somewhere, I need to be able to say, oh, I think this could be an opportunity for my Data science team. I can define a problem. I feel like everyone should be able to do that. But just like you don't know how, most of us don't understand how our engineers work in our cars like it is a mindset, right? Like actually jump to another meeting that's going to see some celebrities in here doing nothing, nothing but celebrities.

Harpreet: [00:42:35] Oh, jeez, man. Celebrity.

Speaker6: [00:42:37] I know I can depend on this, too. I think that there's a Data literacy is, I think, going to be more important for damn sure. Like, but being able to formulate a question and a hypothesis and Data to solve it. But that also means going to be working hand in hand with people who can probably implement a way to solve it. But yeah, I see Data literacy becoming more and more key. I don't think it's a skill that you could be without if you're in any sort of business role, unless you're like a carpenter or something and it doesn't matter.

Harpreet: [00:43:07] So what about when it comes to ask the question here about or you mentioned about being able to come up with a question, then talk about scoping a problem. I'm wondering what are some kind of, I guess, framework, for lack of a better word that you guys have when it comes time to formulate a question or scope of problem? Then what about you? Then after that, we'll go to whoever else, me, Greg.

Speaker3: [00:43:34] So you're talking about scoping a problem or are we on the other question?

Harpreet: [00:43:37] Well, yeah, I guess the school

Speaker4: [00:43:40] Before we move on to scoping, even if I say one quick thing on the other, the other question in terms of skills for the future, I think that problem solving or the ability to to identify problems and also start getting to the solution, kind of piggybacking what Ben said is always going to be important. The way we do that is probably going to change in the foreseeable [00:44:00] future, the tools that we use, whatever that might be. But being able to getting to the next question creates some frameworks are use some existing frameworks that allow you to do that more effectively is what a service economy is going to be about. So if we're moving away from manufacturing, we're moving away from whatever it might be in the US teaching frameworks for thinking, potentially making tools easier to digest or whatever it might be. Those are going to be the skills that are important to learn. As someone who sees a lot of new tools, a lot of new things and Data science, the most important skill that I see for Data scientists to have is to pick things up quickly. So if I see a new library that's out there and I'm like, wow, I want to use that, I know I can probably learn not in a couple hours if I need to. And same thing goes for anyone in a skill skilled labor type of position is that, hey, I know I'm going to be I have to solve this problem. I can use X, Y, Z tools to solve this. Can I pick up those tools because there's an abundance of them and be able to make them useful to me.

Harpreet: [00:45:01] Absolutely. Learning how to learn is a skill and a superpower, which is why you guys should tune in to the episode, are released today with more Brockley. We talk about learning how to learn. Didn't feel great. Let's say it. Let's hear from you. And then also then on on actually a combination of both questions. And the first question I asked is, will everybody need to be, quote unquote, a data scientist or possess Data science skills? And also, if you can talk about problems, coping requestion definition as well either or.

Antônio: [00:45:30] Yeah, I wanted to build on what Ken was saying to everybody else about is did I say is that why people don't tend to forget is that I don't think you can replace the ability to run experiment, to gather more data. You will find situations where you don't have enough data on hand to answer what is the best way to solve a problem. And for that, you have to run, experiment and capture data [00:46:00] to to to, uh, take action. And that data is critical for knowing what kind of model you need to build to go after the solution. You're trying to you're trying to implement. And these involves especially data scientist to. Design those tests to test, to support the right questions and run these experiment and capture the raw data for that. So when I hear is Data leaving is Data scientist. That's not true. How how you're going to automate that when you cases are so different and so unique that that's why experimental design exists in each of these experiments need to be customized to the solution you're trying to bring together. And for that you need experts for this. And this is not going away. And it's a powerful thing within the Data science community, especially when they partner with behavioral economics experts, marketing experts, you name it. How do I know how many, for example, marketing emails I should send to my customers to boost revenue? Do I know if these three, five, four, how do I run an experiment for that? How do I capture that data to see whether it's successful or not? And based on that data, how do I build a model that tells me how many what's the optimal amount of subscriptions or emails to send to my subscribers to to boost revenue? Those are the things that those are questions that cannot be automated.

Antônio: [00:47:36] And they need to be they need humans to to test in terms of scoping. It's a it's a long process through my experience of it has taken me three months plus to even school. And what I mean by scope is defining the problem, aligning [00:48:00] with stakeholders and coming up with a high level solution. So a high level design for that solution. So it takes an effort between the tech folks in the business folks. So the business folks won't have to agree that, yes, this problem cannot be solved by anecdotes. Are you sick? So you can't be something like or we know if they give you a quick example, if they agreed to on board our tool after the what do you call a free trial, most likely they will purchase more products from us. Right. So this is something that doesn't mean any computer. You want to make sure that the problems you're tackling, it cannot be answered easily just by a simple question. You cannot be answered by a human. It needs discovering patterns in Data once you are able to convince business folks that, yes, you cannot just bring your intuition into this, it needs more than that, then the business can move on to another thing, which is what is at hand.

Antônio: [00:49:23] Who do I need help from? We go to OK, we need help from Data scientist who can help us answer those questions. We cannot answer just by looking at Data with our AIs. This is where you go into now involving data scientist for the scoping piece. In terms of what you experiencing in terms of pinpoint what is the data that you have is a viable solution. Do you have the right data already available to run some analysis, perform some basic basic statistics and see where you are and then start thinking about what is the endpoint, [00:50:00] what is the outcome that you want from this? And now with this outcome, what is the action as a business stakeholder you need to take based on that output? And that's where a lot of people are stuck. We can build models very easily. And then what do we do with it? We have a model that makes many different inferences, that makes many predictions. And Aroostook, with those values and we can take any actions or we can take actions fast enough so that scoping out the business problem, understanding you need to already have an idea of what needs to happen before you go into building. And that's the framework I usually use

Harpreet: [00:50:43] Like that a lot. Is there like a subtle difference? There's no difference between asking a question and then kind of scoping out the question because I feel like asking questions kind of just colliding ideas, creative. But then once I start scoping the question or scoping the problem, that's more active, you start really sort of designing type of experiment or something like that. That makes sense.

Antônio: [00:51:09] Yeah. The funny thing is the the business folks most likely will come with a problem. The questions are typically coming from the scientist. It is that makes. Is this really true? Can you give me a proof of that? Can you show me where this is coming from? How is that Data populated right there? How do you collect this piece of data? Because they want to remove themselves from the hole into it, as I've seen this for five years. I know it's true type of mentality that typically the business folks will bring and they want to be as neutral as possible and listen to what the Data is saying. And then when you go into the whole business, Deep Dove, it's another painful one, because now as a business guy, you want to make sure you go through the whole process [00:52:00] with your data scientist so they can understand the origination of these events that create Data. And that data is going to be used to find a solution. So funny enough, those questions are coming mostly from the technical side, then the business folks themselves.

Harpreet: [00:52:17] Then that's let's hear from you. And by the way, if anybody has any questions or comments and wants to add to this particular topic, please do let me know actually to the queue. Everyone's voice is welcome here. Go for it then.

Speaker3: [00:52:29] So when it comes to I'm going to answer this backwards, when it comes to scoping, you've got sort of these extremes and everybody in business is on someplace on this scale between I have a data scientist that I have no idea what to do with them, to have an organization that's ready to do a lot of the more rigorous activities involved in scoping correctly. And so you have to, at the very beginning, figure out from a maturity standpoint, where is your business? More than I'd say. More than half of Data scientists were hired to tell the business what to do with the data scientist. I would say it's probably about half. I don't have the real don't ask me for data. I don't have it. I'm kind of making that up. So I would say about half and I'm going I'm tripling down on it. Yes, no Data. But you're looking at a process and Greg's got like that's the endpoint. That's that's getting towards as good as you're going to get as far as a process is concerned. But you're often dealing with organizations that are still trying to figure out what the word digital actually means when it comes to their life. And you have people who have not touched a computer as part of their standard workflow and they're not used to interacting with that digital piece. And so you're not getting any data at all.

Speaker3: [00:53:59] I mean, that person [00:54:00] might as well be working on in the inside of a black hole. We have no idea what this person is doing. And there's so much of this around, especially bigger businesses that are on that very, very immature side of scoping. And so step one, really, in trying to figure out what you should do as a data scientist is figuring out what level of maturity the business is and how much of the business can you actually access with data. How much of the business can you even gather data about? How much of the business then? I mean, this is funny, but it's true. How much of the business are people scared of going and looking at? Because there are business units, especially in companies where there's been multiple acquisitions, there are business units that the C suite is scared of knowing how they work. They continue to create cash and no one wants to know. And so I have to warn everybody that ever goes into SCoPI. If you don't watch out before you walk into some of these rooms, these rooms have minds just all over them. And so part of the exercise of scoping is first things first, go to places where people want you figure out what the maturity level is and then go to places where people want you and then get those people to talk to other business units.

Speaker3: [00:55:12] And instead of you saying, hey, why don't you ask me to come in? Have those people become your advocates and then those advocates, those people that are actually promoting you can get you into rooms that you're not able to get into on your own. And that's how scoping really starts at the Wild, Wild West phase. And so really look at it from the perspective of is it possible to do what Greg says to do because that's where you want to go and you're probably going to find out. No, it's not possible right now, but you have to figure out how to get from where you are and you're just a data scientist. But if you don't do this, your team fails. And I can't emphasize this enough. You don't want to be a strategist. But I'm sorry to tell you, if you're in one of those companies where you're [00:56:00] not even digital transformed yet, you are your strategist. I'm sorry. If you don't want to be, it's time to go work at a very mature company where most of this stuff's already been built out because otherwise it's you. Sorry, it is you who's going to have to do all of this. Look, I said I kind of answered. First question and then backed my way into the second one, but hopefully that made sense,

Harpreet: [00:56:24] You should have been speaking directly to me because this is the point of our conversation we had last week. It's not easy. Yeah, I can go for it.

Speaker4: [00:56:34] I actually have a question that comes off the back of that and I have some thoughts on that, which I will also share. I don't want to frame the discussion too much, but what is the best way to integrate or what are some unique ways to instigate organizational change in that sense? I found something that has been really powerful but difficult to harness is morality. So if you're building prototypes or building tools and you're sharing them internally, these things become shared across your organization and eventually they reach the top and literally everyone's already using it. And so you're kind of bottom up approach to making change. Unfortunately, that's a way that this has happened a lot of the time, because at the top of a lot of organizations, the CEO, CTO, whoever it might be, is generally can be clueless about what's going on, hands on on the Data teams. And I was wondering if if people believe that that's a viable long term approach, the morality aspect, building prototypes and tools and getting people to use these things even before they're they're necessarily called for or they've a plan around. These things have been put together, project scope, whatever that might be, or is there a better way to get through to organizations where at the top they just don't get it? Is there another way to force their hand

Harpreet: [00:57:53] In, Tony? I feel like you might have some good insight here, what you think.

Antônio: [00:57:58] So I think I try kind of [00:58:00] both approaches where we used to say, well, I'm going to put it in front of you and you're going to see how cool it is, and then you're going to be going to start using it. And I think it does work up to a certain point because they like the prototype. They're like, oh, yeah, that's great. A lot of times it's happened and they're like, we love it. But if it didn't solve a lot of times, I guess differentiate between them saying I love it. It's the difference has to be it's nice to have versus this is a business critical because a lot of times are creative projects and they're like, oh, I love this. And after, like two months, nobody's used the. Well, I you say you love that and they're like, well, we do love it. But right now we're focused on something else that's not really our priority. And I think that's what has worked for me, because we kind of started an organization and we were trying to really breaking into these different themes. And the best advice I got was you have to kinda do the dirty work, but they don't want to do so. And we will go into the room and it's like, OK, what is it like? It's boring or it's dirty Data or they don't want to clean and doesn't much.

Antônio: [00:59:19] Those are maybe those have provide too much value but it's kind of they don't want to do that. So it was like, all right, this is where you don't want to do, I'll do it for you and then kind of to get on their good side. And then eventually, once you get on their good side, you start instituting those changes, kind of like what you were saying before. When you make your videos, you kind of like nudge minister in direction slowly. So you deliver the same thing it's been. I'm going to give you exactly what you want, even though I think it's not the best solution right now. I'll give that to you and then slowly I'll sergia changes. So hopefully that makes sense. Like if if you asked me for a machine learning [01:00:00] model and I don't think is the best use case, but it's not do much work on my end. Maybe I'll create that for you, then I'll try to nudge you into a little bit of a different direction to show you how it could be better.

Speaker4: [01:00:11] Is there some value from what you said? What I what I really liked is creating value for people, saving them time, whatever that might be. That doesn't necessarily clearly equate to business value. But is that perhaps a way to essentially weasel the idea into their head? Is that, hey, like these analytics have have created value for me with my time, with my specific job, is that even more powerful sometimes than than creating millions of dollars for your business? If you save five other people on an adjacent business unit five hours a week because of something you made, and it makes the business in theory that makes the business some money, but it's not exactly quantifiable. Does that goodwill carry through or is that just like, hey, we we made it and and it didn't make the business money. And so we're going to been in these types of things.

Antônio: [01:01:02] Question? Yeah, I think it's weird because ultimately you're dealing with people and when so when I was back in like hands on, I was working with non-technical people and I automated this one woman before. She was kind of like a financial analyst. And it was see, she went. Forty hours of analysis to kind of like 15 minutes, up past 100 percent automated and wild Data wasn't that transformational value, right? I basically saved her salary, I guess, but I don't know how much she was making. But then after that, the weird thing would have been, like you're saying, I don't know if it's a will or something like she will be in another meeting with somebody and they would need some kind of analytics. And they don't have an answer like, oh, and Tony did some Data last time, in a word. Well, let's ask him again. Or it was just and it kind of works for the business as well. I mean, I so I work at Verizon, so to move the needle [01:02:00] like one percent, it's like a crazy amount of money. But also I think for yourself, it creates real will because ultimately, whatever happens going to be Verizon is going to be OK with me helping out those people.

Antônio: [01:02:11] Kind of like you're saying, you create value, but you also help yourself. You help your career is kind of like gets people talking. And then when later when I ended up switching teams, what I found out was that a lot of the people love working for me with me as a data scientist for them or like data analyst. Was it because I made them feel safe, like every time they asked me something, hey, can we do this? Of course we can. Let's go. Let's do it all get done for you. And they kind of felt very relieved. So like you're saying, those things can be measured because afterwards I'm like, why do you guys say you miss me? Person next person is better than me. Obviously, I like coding and doing the Data stuff. And then afterwards I realize well, I mean jokes. I made them laugh and made them feel safe. And ultimately that's what people are there for. You know, they want to work, they want to improve the company, but they also don't want to be miserable. They want to have some fun.

Harpreet: [01:03:07] And it's an excellent point. Thank you very much, Antonio. Yeah, absolutely. Then let's let's hear from you on this topic.

Speaker3: [01:03:15] I think it's interesting when you look at this from a short term perspective, you've got to get those wins and you've got to get advocates on your side. And that's something that I talk about a lot as you've got to get some sort of momentum started. And if that means that you have to get one key player on your side by writing a report for them that takes you two days worth of work and might make the company twenty five dollars, that's good because it's twenty five dollars a real savings. But you have an advocate. And if you spend a month picking up five or 10 advocates during a year, that's not a wasted month because every single one of those people are going to allow you access to other parts of the organization because you simply don't have the relationships yet. And so [01:04:00] that's part of building out a new organization. And I'm kind of talking down the road. But you are building a data and analytics organization. And again, this is back to whether or not you're doing a lot of this stuff. You're laying the groundwork for a lot of the ideas that are more complex, more more revenue generating. And so that's your beginning. But you also have to remember the C suite is lurking and there the sharks swimming in the amount of money that you cost and you cost a lot. And at some point somebody is going to look at a data science group and say, OK, I need to see revenue. Not so much cost. Somebody is going to say, look, I need to see revenue because you talk about finance as a group.

Speaker3: [01:04:49] Yeah, you need finance, but are they a revenue generator? So they have a really hard time getting budget for big initiatives unless they make it a strategic initiative. And you don't want to have to do this long, long sales pitch for everything that you do. So at some point you have to start generating revenue for the company. And so you have to translate all of that goodwill into a product line or at least into a main feature in the largest product line. You have to start putting revenue around your neck. You have to put that around so you can walk around with, OK, I have advocates and check it out. I have a dollar value, too. And that's really there's a transition point where you have to go from small projects, little quick wins that prove and build relationships to I have cash, I make money. Please don't destroy my team because you're building a data and analytics organization and you're going to have to get justification for infrastructure to do meaningful projects. You're going to have to justify headcount. You're going to eventually take over resources from other organizations, [01:06:00] which no one's going to be happy about. You have to have C suite by and for that, because you have to bring this all under one roof. And so everything that I'm talking about is this sort of walking towards something that is sustainable, revenue generating versus cost center and that. Is now a strategic business unit, not just a team of Data scientists, and so, yeah, you definitely start wherever you find yourself, but realize that you have to quickly move someplace much, much further forward,

Harpreet: [01:06:33] That some knowledge bombs. That's a phenomenal event. Thank you. Would be an awesome podcast with the on campus, but something that a good, good episode to get you to make that happen.

Speaker4: [01:06:43] Some literally just send him the invite.

Harpreet: [01:06:45] So, Greg, go for it.

Antônio: [01:06:48] Yeah, I fully agree with what and what Vince said. I think it's for me the best framework is to have a scrappy solution, a scrappy solution that gets you the quick win that Vin was talking about. And those quick wins, they need to be directly connected to some sort of sales, additional sales that you're generating for the business. With that scrappy solution, you need to find you're a champion. You're a champion is a stakeholder who may be a decision maker or somebody who will speak on your behalf. And that champion also will also you will already talk to that champion about your big vision, the opportunity. What is the big opportunity out there? If you built that Scruby solution into a fully functional product, that person already needs to understand and align with you on that. And then once you have that person, then you're prepared what I call some sort of what if analysis with options that you can present to bigger, larger group of stakeholders. [01:08:00] You always want to leave the decision maker options in that technique. To me, it's about putting the fear of missing out, missing an opportunity.

Antônio: [01:08:12] And in this sense, you have option A that gives you if you go after this, this is the opportunity, this will give you Data already shows you that a scrappy solution is resulting to these added sales to your business. If you give me budget, a budget, I will be able to go after this book right here, which is in the hundreds of billions of dollars. Option B gives you another type of bucket and of some C, potentially it may be the worst case scenario. We're getting lost footage, missing opportunities and getting letting our competitors take over from us. When you do this, you let the decision maker analyze what you have done in a quick short time period to generate value and also what could be in terms of opportunity and also what they would miss it if they don't take action. This is all you need, a champion, a scrappy solution in the framework to help the decision maker make a decision at the fast at a fast pace and give you some budget to go forward.

Harpreet: [01:09:18] This is one of those episodes I'll be listening to again and again. Thank you very much, everybody. Real quick, you had some awesome comments here about defining the question. I'd love to have you unlock it from the chat and share it with us.

Speaker5: [01:09:36] And yeah, so I was more talking in terms of my experience, trying to define a question. And it's helpful, but it's not always easy to do that when you're talking with so many stakeholders. And at the same time, everybody telling you, especially me as a public [01:10:00] servant, I work for the state government. So we are obligated to answer questions that come from the public and everyone is throwing at a million questions that come to you at the same time. So you need to narrow down which ones are you going to answer and will we be able to answer. So that's just, I think, more applicable to my area. I am not sure how I've been. Domain's usually, yes. But I think from a government perspective and there are so many questions that the public ask you when you are obligated to answer all of those, it's definitely difficult to frame that question and then figure it out if you do have the Data. So there is an initial exploring Data that need to happen to some extent before you can actually frame the question and then take it forward through the Data not.

Harpreet: [01:10:56] Thank you very much. Appreciate that. I don't know that a lot of our audience works for the government. That's awesome to have that kind of unique insight there. Thank you so much. Anybody else have any other questions or any other comments we're talking about? Eric, can you comment? So we're talking about earlier. Good. Russell, any comment? I think Russell better for all of us does not look like there any more. Of comments. Thank you guys so much for taking time out of your schedule to hang out. Definitely an awesome session. Probably one of my favorite ones. Number 40, Maska. It's a good one. I'm looking forward to listening to this one again and again. Guys, take care. Make sure you check out the episode that I released earlier today with Barbara Oakley learning how to learn. Don't forget to join in on the Sunday officer session. And if you guys ever show up, why has it been on Sunday mornings? Man, come hang out with me coming out. It'll be fun. And next week, a big episode being released, one that I'm really happy about is with the one and only James Altucher. Um, James has been like such a huge, huge influence [01:12:00] on me over the last year or so. And it's getting him on the podcast was mind blowing. So definitely check that out. Yeah, guys, take care. Have a great, great rest of your evening. Rest of the weekend, hopefully. See you guys on Sunday. And as usual, my friends, remember, you've got one life on this planet. Why not try to do some big.