Happy Hour 98_mixdown.mp3

Harpreet: [00:00:05] But what's up, everybody? Welcome. Welcome to the arts of Data Science. Happy hour. It is Friday, September 30th, 2022. This is data science. Happy hour number 98. Just two more weeks until we get happy hour number 100 in the books. Man. I'm excited for that. Yeah, I'm on Pump man. Number 100. Even though I started doing this thing right about two ish years ago, which makes sense since it's 100. But I'm glad y'all are sticking with me and hanging out. They used to be big. They used to be like overwhelming number of people in the room. They'd be like 50 something people in the rooms. And now it's it's down to a more intimate group, which is all good, more interesting conversations. But if you're out there listening on YouTube, LinkedIn Land, just know that you are welcome to come to the happy hour at any time. That is, just go register bit.ly bit.ly for ads. So this is not number 100. This is number 98, number 100, and we have it in about two weeks. Been a good week, man. I was out in San Jose for the Intel Innovation Conference this week. That was cool. I got a chance to to see the Intel CEO Pat Gelsinger talk and. Debut, a whole bunch of new chips and a new hardware that's coming out. Me personally, not much of a hardware type of guy, so a lot of it was kind of over my head, but it seems like it's about to be pretty interesting what's happening. They're launching their own GPUs. Intel's got a cloud that's going to be popping off, so that's super cool.

Harpreet: [00:01:38] So I'm excited for that and looking for it again. Cheaper GPUs from Intel do some more deep learning stuff. But also at the same conference, I got a chance to see Linus Torvalds, the creator of Git, the creator of the Linux operating system. He had just like a little chat with them with Pat Gelsinger [00:02:00] and then one or rather received the Lifetime Achievement award from Intel Innovation. So that was cool to see. And then I got to see Andrew NG do his thing. He did a presentation on data Center Guy for about 45 or so minutes. I was right there, front row. It was cool to see. It was great, man. Then I got a chance to hang out with Kiko and Mark and Seattle data Guy Benjamin Rogan out in San Francisco for a little bit. Went to this a happy hour to finally meet my friend Ishani as well. Vc are never ventures so that's cool. And then the next day after that with the long ass layover in Vancouver, got a chance to hang out with more friends. John Sebastian, Rory McGill, Zubayr, Mansur, DuPont, U.S. guys all hung out with me. That was cool. Went to the beach and just kind of chilled. It was nice. And now we back here Friday, man. Friday. I'm about to go see their dogs right after this, this session. So we keep it short. Ta ta ta. About an hour max, because a little date night with the wife and go see Verdasco, which I'm pumped for. So I should.

Speaker2: [00:03:03] Be. Oh, yeah, he's good.

Harpreet: [00:03:04] He's good, He's funny. He's funny, man. I like him. I like him a lot. Yeah. So if you're joining in on LinkedIn, let me know if you got any questions. You got any comments? Man, I'm happy to take all your questions and comments. Also, if you're interested and come in live to the session, let me know. I'll send you the link to the to the room and you can join us here in person. Shout out to everybody in the room. Christian, what's going on, love? Need Arch it, Jason. Yalla, Toshi. Good to see Tashi again. Coach Step in the building and kick on the building as well. Some new names, but I'm happy to see all you all here. So anybody got any questions they want to kick off the discussion with? Let me know. Let's let's see this question coming in from a from Christian here. Christian, go for it.

Speaker3: [00:03:48] Yeah. I'm just wondering if there's anything specifically to be aware of during a technical coding panel interview. Panel. So if anyone has any tips besides the content itself, I think we all read enough [00:04:00] about SQL and practicing that and the importance of it to Python if it's relevant for the role. But I'm just more so curious as to like the structure and what to expect. If there's a software system that you log into to do a technical coding panel, I'm just curious to know.

Harpreet: [00:04:16] Yeah, I've seen it done a number of different ways. When I was actually interviewing with Google, it was just a they opened up just like a Google doc and I wasn't able to run the code. So it's all just pure syntax. Other places I've done it was kind of like a I guess coder pad. It's with the software where you just kind of type it out right there on the fly. I don't know. I'm curious to see how else other people have seen it done, but those are the two ways I've seen it, where it is strictly just on a dock, writing out the code text file or whatever, just to see if you kind of understand how to think through the problem and write the syntax. A couple of things to be aware of, man. Like. Apparently you're typing speed matters. That's one of those subtle things that you don't hear about. But just make sure you're a fast typer. And then also usual tip is just think out loud. Just kind of think through and be especially verbose. I think the worst thing you could do in one of these situations is just clam up and not talk or think your way through the process or rather just clam up and just think internally rather than verbalizing because you want to give the interviewer a chance to kind of see what's going on in your mind. Head.

Speaker3: [00:05:22] Yeah, no, I was just, I was just going to say it's almost like explaining, like being verbose, like explaining your thought process. Like what would you say? Like if you get it wrong, but you explain your way of thinking. Like, I know it probably depends on the role in the company and how senior the role is, but how would an interview take that? You know, like you're not you're probably not going to be 100% on every single question, right?

Harpreet: [00:05:51] Yeah. I can only kind of think about personally. Sure. What I would do in a situation if somebody like kind of, you know, [00:06:00] not head in the right direction, not going the right, you know, kind of way, I kind of nudge them towards the better solution instead of instead of just saying, Oh, that's complete wrong and not give him any feedback. I try to actively give feedback and nudge them into the right direction because interviews are nerve wracking, right? You can get nervous. He can kind of have have your brain freeze up. Yeah. And I think for an interviewer who's on the other side of a table, they should be kind of sympathetic towards that and help nudge you in the right direction. But I'd love to hear what other folks are say. Jason, any thoughts on this art? Any thoughts Navneet Any thoughts on coding, interviews and any tips for Christian? He goes in the building as well, so making coffee on the inside. So let me know. I see Michiko is coming on. She's about to drop some knowledge. She's on mute as well.

Speaker2: [00:06:54] There you go. Sorry. Can you see me? Yes. Sorry about the Miss Must appearance. I'm finishing my two week vacation before I hop to my new stage of career. Yeah, I think one way to kind of think about it that sort of helps me, because when it comes to, like, questions, I know when I was doing some of the tech interviews sometimes like I felt like, Oh, should I be asking like questions about X, Y, Z because maybe I need this information, but I also don't want to seem like I'm stupid. So I kind of treat it as if like you literally have like a business partner, depending on who's interviewing you either like you're like business partner is like right next to you or you've got like another like analyst on your team. And then the normal questions that you would ask them as you're going through the analysis or through the screen, that's how like I would sort of think about it because like so for example, if you got like a tape, like if you got a sample table, right, and they're like, create a report or what have you. Right. The normal questions you'd be asking is like, [00:08:00] okay, well, first off, for these metrics, like how should I be like, how should I be calculating them? This is how I think. Like, for example, if you're if they're asking for like an annual recurring revenue, right? It's like, okay, when I see this, like, this is how I think about calculating.

Speaker2: [00:08:15] I sum up all like the order values and all that. How would you, how would you define like annual recurring revenue, right, for this problem? So like asking for business logic I think is always like and clarifying that I think is always really good. The other second part too, is you can't ask them for solutions. But there are studies that do show that like interviewers, when they talk about themselves, they like you more, regardless of whether or not the content that you are producing or like the quality of your output. So that can be we're also asking questions can be very helpful or even making comments like, yeah, in my job as X, y, Z as in my job as a sales analyst, you know, I used to do these kinds of reports all the time and these were some things that popped up. Should I go ahead and filter out for like null values or what kind of level of cleanliness do you want me to go after? Right. So it's things like, like that's how the mindset I would sort of do is like if it was just your teammate or your business partner, like, say, in front of you, what's like the normal dialog? You would have to kind of get to that resolution.

Speaker3: [00:09:29] It's really helpful. Yeah, yeah. That context and like even how to interact in that, I've never been through it, so that's why I'm asking. But yeah.

Speaker2: [00:09:40] Yeah, it's, it's almost like, like talking to a client, right? Sure. With a lot of the interviewers sometimes. There's a couple of different ways they could kind of get on the panel. If it's the hiring manager, it's because they want to hire for their team sometimes to they'll what a company will do is they'll just basically say like, hey, [00:10:00] like every every engineer or analyst has to do X amount of hours like interviewing. So if they're having like a big, like expand like head count, for example, like, oh, we want to double the company, right? They'll just say like, Hey, everyone, you have to do like 5 to 10 hours, like regardless of how senior you are. And so some interviewers are going to feel like awkward ducks. I know someone had said, like, if an engineer doesn't sound like a walrus when they're giving a presentation, they're probably not a real engineer. So that's almost like how to think about it. Kind of like some interviewers are going to be nervous because they're really new and bad at interviewing other folks. They're like doing some kind of rotation. And then they might they might look at your resume, but probably not honestly sure. So it's like you can kind of just start the conversation as if they sort of don't really know a whole lot about you.

Speaker3: [00:10:54] Mm hmm. Yeah, like I've had technical interviews, but it was more discussion or like framing a technical concept in a like it's like a normal, non-technical business stakeholder.

Speaker2: [00:11:07] Yeah.

Speaker3: [00:11:08] It's like in the context of SQL and look em out, but never actually like sitting online and sharing a screen and coding live. I haven't, I haven't had that yet. I've done like a skills assessment test, but I got to take that home. It was a take home project and then present it. So I had more time, you know, to not in real time. So yeah.

Speaker2: [00:11:27] Yeah. And I think the other advice I give is as structured as you can be is, is great. So for example, if you want to like already like on the Google doc or coder pad, say like okay, here's the, here's like the key terms and like the values that I need to be calculating or the metrics here is the current definition. As I understand it, this is like the time period, all that, and then you just kind of like code out the query and just ask them like, does like this. There's like the business logic kind of like make sense or are there any questions you [00:12:00] have? Or and then usually when you say like, are there any questions, they might say like, oh, but what if like there is like different currencies or how would you handle like duplicate orders? And then what you can do is you can take that original one and then just like underneath it, add like the change that you would make so you can kind of approach it in sort of like the same logical way that you would a client. Cool.

Harpreet: [00:12:27] Some great Starry night questions with great comments coming in from LinkedIn, from J. Michael Palmeiro, the fourth shout out to J. Michael. He's a director and head of developer relations at at Telmex. He says, Be true to your thought process if you're talking through your thoughts, even if you're unsure. Keep up the verbosity and positivity. Approach it like a fun challenge. The ones I have seen not do well here. Freeze and look panicked. I think that's great advice, man. Just have a good time with it. At the end of the day, like, you know, if you're coding for a living, you should enjoy it. And so show that you enjoy it in the in the interview as well. I'd like to holler at Eric, get his perspective, because I know you're actively just have been interviewing for these live coding sessions. And so Christian's question was anything to be specifically aware of during a technical coding interview panel besides content and self, any tips, handling, etc.? And this is one of those situations where it's like happening like live and in like in sync. Yeah. And if, you know, Jason asked if you guys got any tips for for Christian here just use the raise hand icon let me know I'll I'll get to you guys.

Speaker4: [00:13:42] So one of the things that I think this is just like a little practical thing that I think makes things easier is when you're working in an environment like code pad or whatever, it's really easy to, you know, you're going to write probably not a really complex query. [00:14:00] And so it's really easy to just kind of like write everything on a couple of lines. But like, I actually think that that's kind of a little bit of a disservice to yourself because I'm not necessarily like great and hard core on whether or not like legibility is great, but if you're working through something and then you kind of get stuck and you got to go back and work through something, rework it. It's nice if it kind of already flows the way that you usually think. And it's also easier to just like comment something out and try something new and things like that. So I think that just a tiny little practical thing is. Keeping things a little bit clean as you work helps keep, I think, keep my thoughts organized as I'm working, especially in an environment where it's like stressful. We feel timed. We don't know whether or not we can Google things and we're talking to somebody we've never met before, you know? So it's like anything that you can do to make it familiar to you and like, keep it at your speed and your jam is great.

Speaker4: [00:14:55] I would definitely support what I think was Harpreet or Kiko saying of walking through your thought process. I really appreciate it when people do that, even if even if it's just kind of saying, Well, I think here would be I think it'd be helpful to use a rank here, but I guess you could also use a row number for this in this case because like, you don't have to use both of those. But I can now hear you say things if it's a sequel thing, for example, to be able to say, oh, like, you know, it's both of those or I'm not necessarily going to be like, Well, when would you use one and when would you use the other? Because I don't really care. Just, you know, you'd figure it out when you need it. But I really value people being conversational and talking through their thought process because the other thing is you would be my coworker if you get through this interview. And so I would really like to hear from what you hear, what you'd be like as my coworker, because I want to hire cool people. So those are kind of some thoughts that come to mind quickly.

Speaker3: [00:15:54] Yeah. Thank you so much.

Harpreet: [00:15:57] Let's go to Navneet and then co sub and [00:16:00] then I'd love to hear from like Arch or Jason man, just because you know y'all are new round here, man. I want to see what you all sound like. Let's go to Navneet CO sub then RJ and Jason, if you guys want to chime in, please let me know. No pressure.

Speaker2: [00:16:14] Yeah, I mean, depending on who's interviewing you and what their style is, you know? I think it's a big factor. It's a junior person somewhat. You know, you may be kind of bogged down a little bit into the weeds. But if there's somebody a little bit more seasoned, you know, they'd like if they like to probably know your process from start to finish, even though you can't get through the entire code. I always whenever I have these interviews or I'm taking these interviews, I usually ask the candidate to get a project that they're comfortable sharing. And so that helps me know what their process was from start to finish and do like a screen share or something. So I like to keep it handy. If you had the opportunity to go there and that way you sort of see the sophistication of thought and your code as well. But yeah, I mean, all really good advice, advice there. And again, like ask questions because sometimes I've been in these interviews and they're like really vague, so it's totally fine to take your time to ask questions and get clarity on what the actual problem is. Sometimes they are meant to be vague, so you ask them those questions.

Harpreet: [00:17:43] Whatever you do, don't don't give up in the in the middle of the interview. I've been in.

Speaker3: [00:17:47] Situations.

Harpreet: [00:17:49] I've been in situations like I like freeze up and I'm just like, Oh, shit, I can't think through it. And and that happened to me during the Google interview. I was like, throwing in the towel like mad. I can't do this. And [00:18:00] the guy was, like, super helpful. He said, No, let's give it the old college try, man. Let's. Let's think through this together. That's awesome. And that was just really like believing. I was like, Here it is. This guy's like, You know, I'm interviewing for Google. I think it's like one of the hardest things ever. But he's like, super nice about it. And we get through and get a solution and, you know, that was helpful. Kosta was Go to you, then arch it. And then by the way, all you all watch on LinkedIn. Super excited to have you guys here. Give me one favorite. If you're on LinkedIn, give me a smash like, you know, share this thing, get get the word out there. And if you have a question on LinkedIn, got a comment. I always welcome all of your questions and comments, so please do keep them coming up and then it.

Speaker3: [00:18:39] So there's a practical aspect to interviewing, right? And it's going to change. Your mentality is going to change, depending on whether to be blunt, whether you're in a position of strength or not. Right. Are you coming at it from a, Hey, I have a job that's pretty good, pays me what I need to be paid, you know, and I'm looking for something different, or am I coming at it from a, Hey, I'm trying to break into this industry. I'm trying to secure trust that I don't have the experience for and things like that. Right? It's a slightly different mindset between both. And obviously, it also depends on what's the competition like in that in that market, Right? The competition in the American market, particularly for junior entry level data scientists, is fierce. Extremely fierce. Right. Same thing in Australia. But at the moment you're talking about people with 5 to 7 years experience. Very different ballgame, right? Typically, people interviewing at those levels don't have as much competition because that's the current state of the market that it is. And the second aspect of it is that typically you're coming in from a different from a different role either in data science or maybe not in data science. That is still pretty much enjoy, especially if you're entering from a side from a different, sorry, entering from a parallel kind of field, right? Now, I've been fortunate enough to mostly [00:20:00] interview from a position of strength.

Speaker3: [00:20:02] So what I tend to do there is I tend to focus on being myself and where I to almost quote Stevie Ray Vaughan. Where I mess up is when I start thinking about what's the technically correct way to play up and down the fretboard. Right. But I'm just feeling it and just going with my natural self. It just works, right? Like that holds true for so many things. Like I did an interview recently and it was a panel interview where three people were watching me code and my ID and it was the simplest Python programing thing and I was just messing it up. I was just messing it up. And and there is no way to really explain that at all. That's something that I could do. Hands tied behind my back. If you just give it to me, leave me alone in the room for 20 minutes or 30 minutes and just smash it out. Right? No problem. So a lot of the time we get so caught up in, Hey, I have to behave in this way, answer it this way. There is different expectations for, oh, you know, whatever it is, right? Should I be using a star approach versus some other approach to request? And should I be focusing on typing speed or should I? Am I allowed to slow things down and actually ask and understand and ask inquisitive questions? Or are they expecting me to just dive right into it, Fire off ten different answers, like we've heard three or four great strategies here, and some of them actually conflict with each other, right? Like typing speed matters.

Speaker3: [00:21:24] Yes. But then do you sacrifice typing speed to to stop back and ask them all the right questions or are they expecting you to jump in? And different interviewers are looking for different things. There is no point really trying to find out what are they looking for? I'll give them that. At the end of the day, that's not a good representation of yourself. Now, I caught myself thinking, Hey, I'm going to dive in and give them the code that they want really fast. Like I was reflecting on that and I'm like, That is just not how I operate. I should have pulled back by the end of that interview. I actually managed to pull that back and say, okay, hang on, let's [00:22:00] pull this back and start from a conceptual approach. I've never finished the challenge, right? I've never finished it, but I managed to show them what my thinking would be in terms of solving the algorithm. And then in terms of how I would then go to code it, it's a totally different ball game, right? And it's kind of the difference between a take home exam and a live coding panel.

Speaker3: [00:22:20] But then there's the other side is take home exams are usually a lot more time intensive. Is it worth investing your time in that? So when I'm interviewing someone, when I'm on the other side, I don't so much care about whether they get to the answer or not. And you'll find a number of good interviewers might say the same thing, because I've heard this a few times. I think. How do they get to the full correct answer or not? What I want to see is how they think and how they approach the problem. Right. So personally, I do better at whiteboard interviews and unfortunately, we don't get to do always. We don't always get to do a take home exam and a whiteboard interview and a live coding challenge to see where a candidate fits best. Right. So if you're a fish out of your water, call it out. Say, Hey, I'm not used to this. I'm much better on a whiteboard. Do you have something you'd like to ask me on a whiteboard? There's nothing wrong with like like flipping the tables or asking for something different. But that's easier to do when you're negotiating from a position of strength, when you've already got a job that you like and you're not, you know, trying to break into an industry that can sometimes add like can, it's essentially about positions that strengthen negotiating.

Speaker3: [00:23:31] Yeah, right. But when you are in a position of strength, ask for us. For what? I mean, it's a bit ballsy, but ask for ask for the things that will show you in the best light for them to realize what they're going to be working with, because that's what they want to see and that's what you want to show. Like you said, you're going to be working next to me. I want to know how you think and how you do your best work, not how you flounder in something that I expect everybody to pass. Right. So it's just a different mentality and mindset. It comes with its own [00:24:00] risks. Be warned. Right. Some employers would hate that. You throw that at them and they would just be like, No. You've got to do things that way and fair enough, right? Power to them. That's their position of strength, where they have enough candidates that they will get enough people playing the game their way. Right. So it's at the end of the day, it comes down to negotiation strength in a way that really helps. Yeah. Provides which end of the spectrum I'm on to. So yeah thank you. I'm also depends how Machiavellian you want to be about the whole situation so sure play cards will you like to.

Harpreet: [00:24:38] For those not in the Zoom room reading the chat, Eric says When I do interviews, the goal isn't to finish all the questions is to do a good job. Asking questions shows good qualities that just writing code may not show. And then he's also saying, Hey, am I not? I'm not used to coding on the spot with unfamiliar data in front of strangers. Is that what I'm going to be doing in this role? That's a fair question, man.

Speaker3: [00:24:59] You know, isn't that the really poor assessment of writing good code? Like we want to see people writing good code, but we also want to see them to do it in 20 minutes with other people watching over their shoulder. I've never seen genuinely good code written that way unless they already knew what the answer should be. I don't know. I'm just. Yeah, I guess.

Harpreet: [00:25:20] So. What an arch it has hand up. So let's go to our chip. But before we do that, there's a good question. That's a big question. Rather coming in on LinkedIn from Schuyler Bullard. And I want you guys to noodle on this while we go through a budget answer here. Schuyler is asking, how does data science make a profound impact in business? That is a massive, huge question. So if you want to start queuing up your answers for that, use the raise hand function. I'll be sure to get to you. But RJ, go for it.

Speaker3: [00:25:48] Now, I'm actually going to take the stage here and answer both the questions in one go, because I really like the second question. But the first question goes, this is something from my personal experience, and I'm going to be speaking [00:26:00] a little bit on the technical side of things. I brought into data science three and a half years ago, coming from a software development background. It was really difficult, but something that helped me was talking about big complexities whenever I was coding, even if it's not required or if I was putting off putting up a for loop, another for loop, I was just saying it out loud. Hey, it's in scary complexity because there are two loops. So at least the interviewer, like some of the people said here, realizes the fact that I understand what it is and good code is important for me equally. Second, I want to build up on what O'Neill said was Whenever you're walking, walking, when you're talking to someone and kind of walking through your project, you can always start with the architect, you know, like you don't need to be completely right about it, but just give hints in your conversation around the architect of the solution. Like, Hey, we use Spark here. We know we had an ideal developer or a plumber who had us with these pipelines. They will stream to this that way. You know, the interview will get a lot of things without actually asking you those specific questions. And you can also through here play through to your strengths. If you can actually take the conversation or dig the interview towards the direction, which are your you know, which line your string theory as basically.

Speaker3: [00:27:36] Yeah, as one of the second questions. The second question goes business impact. I was very glad I'm I was really blessed to work with Duke Energy as my first data science job and some of the business impact that I and my team brought in while working there is very relevant, very present. So I was in the renewable the [00:28:00] commercial renewable business. And so wind, solar joined fleets too. So we used to do data science around basically saving the tax dollars or what is it called exactly, subsidized tax dollars so that we can pull that money or that money back into renewable business. And we would do this by actually flying drones above solar fields, which would take infrared imagers, and we would do computer vision on them to identify bone strings. Another project that I worked on was using anomaly detection. I used SB Prophet Lshtm. A couple of things to basically determine if there is a windmill that's going to fail in another 6 to 12 months. And if you can predict that, you can actually salvage it, fix it, and save close to $350,000 because that would last another 3 to 4 years. Otherwise it'll just burn out and you'd have to replace the whole thing. So the business impact is huge. And some of these, I would say utility and health care, especially because these are some of the industries or the areas which are sort of behind they're trying to pick up and come up. But yeah, I hope I did just justice to that question.

Harpreet: [00:29:27] Yeah, absolutely. I appreciate your answer and hopefully this kicks off a good discussion here. The question that Skyler wants to know is how does data science make a profound impact in business? Any thoughts from a Navneet or Jason or a. At Mickey. Go, Eric. Anyone need go for it.

Speaker2: [00:29:47] All right, That's a loaded question. I mean, I don't know. Where do I start? You know, depending on what you do. A little bit of background. I work a lot. [00:30:00] And on the media side, on media analytics and data sciences, and on a day to day, week to week, month to month basis and I know Eric does similar things. So, you know, we make decisions on how to spend money on a TV ad or Facebook ad or Google ad, things that you're seeing on screen. And we use models day in, day out to see what the ROIs are, What's the what are the response for what you know, what's our top limit? What's where are we spending too much? Where are we spending too little things like that. So on a day to day, I mean, when we are talking about millions and millions of dollars of spend, some people do care about what the ROI is on that. So a lot of this this work is built on data science models and memes and whatnot. So so the agencies and the you know, the ad world runs on these model models pretty heavily know the way we buy media, the way we measure media, the way we sell media. So a big part of what we do and understanding these basics of how all of this mechanism works is important to our roles as data scientists. So yeah, it's yeah, I mean, all the I remember you when you were in school and in economics and you read about like linear optimization and constraint optimization and those things in Econ 201301 classes, you know, that stuff is still being used today for a lot of the work that we do. So, you know, we're going to it's handy. It's handy. And at least I look at those response curves at least [00:32:00] once a week. Not or.

Harpreet: [00:32:04] Cho recent the building Chun over there at Sydney Airport. I got to see some that see some pictures of Jo Reese hanging out with Danny and Kosta. I thought that was cool. That's cool. You all got to hang out. Jo, how long is your flight about to go on? 27 hours.

Speaker3: [00:32:25] Oh, Jesus. There you go. 14 hours back to LAX and then. Who knows? Going to be a long day. Brutal. But it's like it's yeah, it's Saturday here and then I get back on Saturday.

Harpreet: [00:32:38] It's like time travel.

Speaker3: [00:32:40] Yeah. So you're saying that 14 hour flights are not normal to the rest of you guys? Like it takes me 14 hours to get point blank anywhere other than Singapore.

Harpreet: [00:32:51] Yeah, definitely not normal, man. Australia is far. Australia is really far away. Joe But the question maybe you want to just, you know, get a quick answer before you board the flight. It's how does data science make a profound impact in business? Profound impact.

Speaker3: [00:33:10] Um, it's a really good question. I would say that it'll help. And from what I've seen, data science, when used properly, will help amplify, I would say, existing processes to make them better. So things that you need to do at scale will help amplify those. Or so if you choose the right use case and you have the right processes and systems down to do data science, so do pretty well. Sure, that's pretty noisy myself.

Harpreet: [00:33:39] Yeah. No worries. Erik or Makiko or Jason. Any thoughts here? You know, maybe. I want to just kind of rephrase the question. So how about how have you seen data science make a profound impact in your business, whether that's a current [00:34:00] job or a previous job? And let's let's just erase the word profound from it. How have you seen data science make an impact in your business? So, Eric, let's hear from you and then, you know, I'd love to hear from Jason if you want. Or Toshi or Jacob or anyone. Really, man. Let me know. Or Christian let me know.

Speaker4: [00:34:19] So I work as an analyst and so I'm not I'm jealous of like art projects, like flying drones over windmills and stuff like that. I think that's a way cool. But working, working as an analyst, I was trying to think of like, okay, like any any of us could like point to maybe projects that we've done or, you know, projects that we've read about of like tremendous savings by implementing something like linear regression even or whatever. Right. Some kind of a model. But one, one other piece that I want to bring. Or data science or just and just data literacy. Having an impact on the business is like a piece of my job that is not in my job description, but that is really important to what I do is I work with like my main marketing stakeholder, like she'll send me a message and say like, Hey, can we see such and such cut broken down by such and such? I'm like, Yeah, you can totally see that in this data mart and like point her right to it and it's like, Oh, I am now enabled as a marketing person to like build this myself. And I didn't even know it was there. So like in that way, teaching and use like understanding how to use data with the mind and heart of a teacher can have a profound impact on the business because she will know how to do that long after I'm gone, and as long as I'm helping other people to learn and do those things, then now my stakeholders aren't asking me for something that they can do themselves, which frees up my time to work on things that they can't do themselves. And so that's what I've been able to then focus on like projects that I would much rather be doing [00:36:00] that are more, more connected, more to my skills versus their skills. So anyway, have a profound impact on the business teaching and data literacy. That's a that's a big thing.

Harpreet: [00:36:12] Yeah, I like that a lot. That's that's actually super true, right? You teach somebody to like, you know, you can feed somebody or teach them how to fish, that type of thing. Right? Awesome. Thank you. So there's one instance that back from my I'm still a data scientist, I guess, but back when I was like a officially titled data scientist where the project I was working on was having to come up with like these custom discount rates, right? So we had this. We were selling manufacturing equipment for HVAC stuff, and we were essentially selling them to a middleman who then go and sell to the public. Right. And this middleman would have like this agreed upon kind of discount factor that that they would sell our products at. But every now and then to get competitive with one of their bids, they wanted to request more of a discount so they'd come through and have a special discount request. And these special discount requests had to go through a manual process where there's high level executive executives like, you know, like VP director level people, like going through spending at least an hour, hour and a half a day reviewing these requests and coming up with the the appropriate multiplier.

Harpreet: [00:37:22] So. I built a model sitting with them, went out to Atlanta for like a week and sat with them to, you know, observe how they go through this process of coming up with the discount factor and just built a model for that, used some boosting models and was able to deploy that to production. And the end result was not only more accurate and stable discount factors, but within the first quarter recouped like three mil top line. It's not a lot given that the company was like doing 750 million in revenue every year, but the 3 million in the quarter man like [00:38:00] that paid my salary, you know, ten times over more than that. So that's that was, I thought a kind of semi profound impact and the opportunities that that opened up like within the company was crazy because now everybody saw the value of like this data science thing and they wanted to invest in an entire data strategy and all that stuff, which I found out was not my cup of tea, which is why I left by end of the year saying something. That might have been a. Ms.. Unmute there.

Speaker2: [00:38:39] Sorry. No, that was my mistake.

Harpreet: [00:38:41] Okay. So another question coming in or if anybody got comments on that, please do let me know. Let me know. Just use the eraser hand function. Matt Macfarlane is asking on LinkedIn. He's got a personal reason for wanting a job that. He wants to donate 15 K to Ukraine for humanitarian aid to put America number one for the. I can't even say that word. So pretty much the biggest motivator for him getting a job and interviewing for DevOps role when his interest line machine learning is that he wants to essentially use the money to donate to Ukraine. And he's wondering, is this a disastrous point to bring up in an interview? How much personal information can we share? I wouldn't bring that up in an interview. That would not be something I'd bring up. That's. Yeah, that's mean. I'd love to give you guys insights on this. Right. Like, do you share your motivations for. Because he's saying that he's going into a DevOps role but from the content he shares from like the stuff that he's been writing about is very clear that he's interested in machine learning. So I guess he's kind of worried, okay, if they bring this up. Can he just tell them the reason why Mikey got your hand up? [00:40:00] So let's let's hear from you.

Speaker2: [00:40:07] I think the thing we sort of forget is that to a certain degree, interviews aren't about us. It's about the company and the job that we're interviewing for. So I, I don't see anything wrong with basically saying like, hey, because like, this comes up sometimes with questions where they're like, you know what? Like what? Where, where do you see yourself in five years? Or what's motivating you or, or all that? Like, I don't I don't see anything wrong with saying something that motivates me to continue growing, to continue delivering value to being the best professional professional I can be in this area is because I do want to I do love engaging in philanthropy and charity. Like, I don't I don't see anything wrong with that. I do kind of feel like. I mean, there's like a couple of things bundled in there, right? In general. I hate to say it, but I do think sometimes there are concerns about if you bring in like financial neediness in a conversation, in an interview. I think sometimes that may be as accurate as it might be. Someone situation that. Might put the interviewer in sort of like an awkward place where they're selling concern about like, you know, they're not thinking about how do I evaluate you as a candidate. They're thinking like, Oh, shoot, if I reject you, is this going to be a liability? All our stuff. So I kind of feel like as long as you're positioning, positioning yourself as the best professional you can be, as someone who is going to deliver value to the business regardless of the title.

Speaker2: [00:41:49] And like, saying that, hey, like, a passion of mine is charity philanthropy, because that is something that is also a passion of mine, too. Right. I don't see anyone seeing anything wrong with that. It's just [00:42:00] once again. Be sure to focus on the value you're delivering to the interviewer versus like yourself. The second part is, I guess, like I do kind of feel like. If I want to rustle up money for charity, I don't know if I would let that dictate the job that I take. To me, like a job or a project or whatnot, is very, very personal. And my future income earning outlook is based off my current job. And if I know I want to do a certain kind of work, even if I'm not great at it immediately, it makes more sense to jump to that work, to continuously get better. And then if you get like, if you get like an annual bonus or if if you get a signing bonus, you can definitely use that for your charitable options. But it's like on an airplane, right? They say, put the mask on yourself and then you put the mask on your first favorite child and then the second favorite child and so on and so forth. You know, on the children who have the highest earning earning potential or that you like the most. Sometimes those are the same, sometimes those are different. But at the end, like, you still have to put the mask on yourself first and then you can and then you can open up your scope of awareness and responsibility.

Speaker2: [00:43:27] Because I can tell you, if you do a job that you hate the that that's going to show up in like three or six months. And then you're just going to hop to another job and then basically people are going to say like, Oh, you're a job hopper. Get it at a you know, And that's going to impact your ability, frankly, to pay it forward a lot more than like your current sort of like just getting a job, just like. You know, for for my I don't know, like to me it's it's a little bit of a confusing kind of decision making process if [00:44:00] you're really, really interested in in donating money to the crisis that is going on. I would say that there are probably more straightforward ways to do so than using your job to do that. And also some companies also, regardless of what role you're working, they do have a charitable matching program. So for example, into it is matter if you're an engineer, if you're in marketing, if you're whatever, they will do some kind of like matching program. So that's another thing is if it's something that you really care about, maybe look instead at whether the company has a matching program and if they have ways to like, you know, meet your charitable contributions, you can say that that's a value that you're looking for in a company. But I just I don't know. I think there's a certain way I would approach the decision making process and how I would frame the conversation.

Harpreet: [00:44:55] Thank you very much, Mickey. Let's go to Kosta and then Eric.

Speaker3: [00:45:00] Yeah. Dangerous waters, my friend. You hit, like, a lot. What? I was going to say. On the. On the head. Right. It's. Does the company care about that? Let's be brutally honest. Is that a priority for the company in an interview? Yeah, as much as I was saying before, you apply your position of strength, you're there to find out what the company has to offer you in terms of opportunities. It's also about what do you have to offer the company, right? The company is investing X amount of dollars into you when they have the opportunity cost of investing that same amount of dollars in someone else. Right. Is that something they really care about? How what's the nature of this company? If they are focused on charitable causes along the way. Brilliant. Maybe they are receptive to that. Bring it up. See what you can, you know, show that that's an additional alignment. If that's your primary focus and goal of doing it. I hate to say it is [00:46:00] as lucky as we are right now in the data industry is far more lucrative ways of making bucket loads of money to, you know, to fund whatever it is you need to fund. So even from that perspective, just yeah, there's enough people in this industry looking at entering this industry from a passion standpoint that you're not necessarily doing yourself a major benefit by putting that out there at an interview stage. It also depends on what stage of the interview you're at.

Speaker3: [00:46:33] The nature of the company that you're at, the nature of the people that you're dealing with, right. Like some companies might be super for it. Other companies are looking for that going, Hey, hang on, am I paying this person an extra $15,000 a year? So that they can do that comfortably. Right. And it's not an unreasonable like thought process. It's again, sounds and I'm on a bit of a cynical streak today, but it sounds cynical, but it's not unreasonable. Right. You've got to put yourself in the company shoes as well. What are they looking for? Where are they looking to invest their money in people that are What are they doing with that money? Right. So, bottom line, some companies can afford to be charitable and they do set it up. And I admire that. Right. And I would definitely join in. And depending on the causes. Right. That's a very personal thing. And sometimes. I prefer not to advertise things like that as much as I do a lot of community work and things like that. Where I do bring it up is. For example, I do a lot of community work with the music community out here in Sydney, in Australia, the particularly the Indian Indian classical music community. Where I'll bring it up is my passion for that leads to time being spent there and time that I expect to spend in those environments.

Speaker3: [00:47:58] So I [00:48:00] draw those time boundaries quite clearly in that I need a certain amount of energy and time to engage in those activities properly. Right? And that's a much more acceptable, acceptable pitch, I find. Right. Even if it means that you need time and energy to do certain fundraising. I know people that have you know, I know someone who I used to work with had a particular art form that they practice. It was a performing arts thing and they had to make sure that they had at least one day a month where they could do a full workshop. So that was part of the agreement with their employer, is that they would have annual leave, but they would also have kind of partially and partially purchased, you know, additional leave one day a month so that they could go and focus on workshops for this art form because they had to do an intensive day at least once in a month. Right, to actually do well and continue to perform them. So there's different ways of bringing that up. I don't know if coming out right and saying, Hey, I want to spend X amount of money every year donating it is necessarily the right move. It does bring up questions that. I'm not necessarily helping you in your favor of getting that role right. So, yeah, just think about how you present that. It's a dangerous game. Just be mindful of it.

Harpreet: [00:49:24] Thank you very much. Eric, you had you had your hand up there and go for it.

Speaker4: [00:49:28] Yeah. So it's been really helpful. I've kind of had a few minutes to think. So one thing I'll just throw out there is similar to what Michael said, like at LendingTree, they have a matching program. It doesn't matter what Lending Tree's politics are, it doesn't matter what CEOs politics are. Whatever I can donate to whatever causes I want. And LendingTree matches up to a certain amount of it, and I use pretty much the whole thing. I think it's great and so I would definitely recommend that. But one thing I was thinking about like to kind of your second. Your second. Question that's a little bit more general [00:50:00] on on LinkedIn. How much can we bring up personal stories outside of technology if we can tie it together? So. One thing that I don't understand very well personally, but wish that I did better is privilege, right? Like I'm a privileged white guy. If I am into some if I have strong feelings about some political topic, I don't necessarily have to share that in an interview. And no one will know if I have fringe beliefs or something. I cannot see your profile picture on LinkedIn because we're not connected and so I don't see them. But if you happen to be a trans black man, then like some personal stories of your life are like much more visible than my personal stories are. So I guess the kind of like, as I think about it is just like when I think about things I want to share, I think about like, if I share this.

Speaker4: [00:50:57] I want to share. I want to feel comfortable sharing something that I own and I'm going to own no matter what. And that if the company that I'm talking to has a big problem with this, then I have to be comfortable knowing that this company is not a good company for me. And then there are other things that I can share that I can decide not to share because it's it's no big deal. I don't care if somebody disagrees with me about one thing or another. And those might be, you know, stories that you don't necessarily have to put out there. But one thing I think is just really important to recognize is like it's easy to for me, you know, to think from a perspective of privilege, of just like are just like, don't talk about that because it might it might not work out for you, but like, we don't all necessarily always have that option. So it's helpful for me to think through it because sometimes I'm on the other side of the table to say, like, what does that what through what lens am I listening to this story? And if I think about it from a different lens, how would my perspective change?

Harpreet: [00:51:58] Eric, thank you very much. Any [00:52:00] input here? Our shit or Jason, If you guys got anything to say, please let me know if just you know you raised your hand. Feature their on Zoom question on YouTube coming in from Towson. So. It's a good question, said. It's kind of a hello world update of science question. What's the difference between data science, machine learning, deep learning and artificial intelligence and NLP? That's a good question. So. I think of it as, like concentric circles, right? That's kind of one way you can you can kind of imagine it. And the way I kind of imagine it might be different from other people. But I think data science is like the entire, you know, encompasses all of that. I think data science encompasses not only machine learning, deep learning, AI, NLP, but also like data governance and analytics and data management and data engineering. Of course, that to me falls all into the data science umbrella. And then within data science you can have machine learning, right? And machine learning. You know, it's just a set of techniques that data scientists use. And then within machine learning go a little bit deeper is I'd say. I miss the artificial intelligence part, I guess. I think data science, artificial intelligence, even though they have significant overlap, when I think of artificial intelligence, I think of more of deep learning. It's kind of more like synonymous to me is is I think deep learning and AI are really synonymous and NLP is just a you know, NLP would would cross both machine learning and deep learning because there's techniques to do NLP that don't require deep learning. And nowadays there's mostly deep learning techniques for NLP, Transformers and the like. Hopefully that clarifies it. I'd love to hear somebody else's conceptualization of this. Or got anything to say there or coast to coast than art [00:54:00] than art yet. And the next question Towson has, if you guys want to just kind of think about it, is do I need to learn data analysis before learning data science? It's a good question. Go for it.

Speaker3: [00:54:13] So I, I tend to draw a slightly different mind map to it. It turns out to be more of a very overly connected tree where I kind of put artificial intelligence at the top, primarily because there are non data driven processes that mimic artificial intelligence in fields outside of software. Right. So when you come at it from a robotics perspective, there are control systems that allow you to make decisions that are not necessarily well, they are technically data driven, but they don't follow the statistical methodologies and the mathematical methodologies that you see that are such a core aspect of data science. So when you look at it from a skills perspective and a technical perspective, someone who's really good at data science may not have the natural knowledge, like the latent knowledge sorry, not natural knowledge, latent knowledge built in to address those aspects of artificial intelligence and vice versa. Right. Someone who's well versed in some of those systems in terms of artificial intelligence, in terms of decision making, machines might not have the might not be conversant to the mathematical level of what you expect from a data scientist. So there's that aspect of it as well.

Speaker3: [00:55:30] So when you start, I start thinking about it as artificial intelligence starting to bring on a. Beyond just the basic, you know, if else kind of decision making systems that we have, how are we augmenting more abstract level decision making, how we how we proceeding to more complex synthesis level decision [00:56:00] making, Right. Things like that that typically we're trying to do things that aren't very black and white anymore and that can come in many different flavors. So I tend to put artificial intelligence at the top of that tree as the broadest possible category. And data science being a series of tools used to bring us closer to artificial intelligence, robotics like path planning, algorithms within robotics being another another path like another set of tools within the space of artificial intelligence. Right? So I think of intelligence more as a broad concept. But then each of these other things, whether it's statistical numerical processes, whether it's Gaussian processes, whether it's data science, I think of them more as techniques and approaches to fulfilling artificial intelligence. But that's also very biased to that point on it.

Harpreet: [00:56:58] That they'll have that very, very well articulated, far better than I could have said that. Thank you. Kosta RJ, do you still want to go? Then we'll go arch it and and Eric Sims. Let's go. Our shit.

Speaker3: [00:57:16] Okay.

Speaker4: [00:57:17] Maybe to. What I am guessing is kind of the spirit of the question as to the difference between all of them. There are lots of good diagrams, picture versions of what Harpreet and Khosla have both said, I would say. If. If you focus on learning to solve business problems using data, then it might end up being that you need Excel or Tableau to solve the problem, or it might end up being that you need an algorithm that you would that you need in [00:58:00] order to solve that problem. And whether you get that from a post that is on towards data science or that is on ML Weekly or artificial intelligence today or whatever the website happens to be called. Ultimately, if we're solving if we're solving problems with data, then, you know, you'll fall into the you'll fall into the tool and you'll find out along the way that sometimes data science, AI, ml and all of that is just kind of strung together like beads on a string because, you know, when I'm talking, sometimes when I'm talking to people, I'll say I'm going to throw out some buzzwords here because I don't know which one you use. Data science, AI, ML, whatever. Now, deep learning, that's a thing. It's a thing.

Speaker4: [00:58:46] Nlp It's a thing you can either be looking at NLP or not looking at NLP, but the other terms can be so broad that I wouldn't worry too much about it. And then I think the second question was like, Do you need to learn data analysis before you learn data science? You set it so. Yeah, that's a hard question, but what I was kind of thinking with it is like, it depends on what I guess, like how do you define analysis? Yeah, like I think that learning to think logically and critically and then learning whatever tools you need along the way, I think kind of like grows into being called analysis. And if you happen to use tools for solving problems like predicting things and then it turns into using an algorithm to predict something, then I think now your analysis just became how to use for machine learning, data science. Ml Whatever. Ai. And so like if you know how to think logically, then you'll get to analytics or whatever else. Data science words. If you don't know how to think logically or critically or analytically, [01:00:00] then the rest of it's garbage because it's all built on being able to think and think in those terms and.

Harpreet: [01:00:09] It wasn't very much. Yeah, go for it. I want to quickly.

Speaker3: [01:00:12] Add one thing there that it doesn't it doesn't really matter if your data science data, if you want to learn data analysis first and become a data scientist, just play to your strengths. If it's just about the tools, it's about the result. More often than not in the industry, it's about the end goal. So Erik is an analyst. He knows a lot of things that I don't like. I don't know. W I don't know. He's a far better skill quarter than I am, but I have been coding in Python for the last six seven years. So that's my choice of tools or that that's a strength I like to play with.

Harpreet: [01:00:58] Thank you very much. Yeah, I mean, I agree with all that. Like, do you need to learn data analysis before learning data science? Well, data analysis is kind of a core part of data science. It is an essential piece to the puzzle. So I think you're you know, as you're learning data science, you will learn data analysis, I think. I think. I guess. I think, yeah, you probably end up learning data analysis before actual data science stuff. I think that progression makes sense, like before you get to like, like, okay, so for example, right in my career now, I'm mostly doing deep learning type of stuff, right? But I started out in grad school studying like, you know, math and stats. And my first job in the field was as an actuary. Actuarial sciences. It's a lot of data analysis and statistical modeling, right? I wasn't doing machine learning so much per say, but we're building statistical models in some way, shape or form. And then it became a biostatistician, which [01:02:00] was mostly all data analysis and zero machine learning. Right. And then moved into other roles in proper data science, and that was more machine learning type of stuff. So, man, I don't know where I was going with that. But yeah, I think inevitably you end up learning data analysis skills prior to learning data science. If by data science you mean machine learning, deep learning and, and things like that. Miki, you got anything to put in here?

Speaker2: [01:02:31] So there are these three stages in a typical mall development pipeline called feature engineering model analysis and then model evaluation. So. Yeah, because if you put together or train a model and it doesn't perform well, like for, for most people what they would say and this was something that people were talking about too when AutoML started coming out was like, AutoML might like switch out the different like model architectures and you can, you can kind of like swap out like a random force regression or like whatever, like a different algorithm, but the power is like in the feature engineering. And more importantly, if you don't have the right data or you don't have strong signal in the data, like guess what you have to analyze and you have to go back and get some more data or you're going have to decide whether or not to even close the project, right? Because that whole statistic about how like 80 or 80% or whatever, malls don't make production, it also includes malls where they trained it realized the results were garbage, and then they then had to throw it out. So I'd say like analysis is like super important. I'd say storytelling is part of that as well. But I think where it's sort of where you start kind of veering off a lot of times is like the strategy and the decision decision making part as well as like interfacing with like key stakeholders, for example, like product or marketing or whatnot, then determine, okay, so, you know, we've [01:04:00] seen that we have this trend or we have these results from our analysis like now what do we do? Right? So as a data scientist, I and then eventually, like as an engineer, I became very removed from my business partners.

Speaker2: [01:04:14] So I no longer had the insight and I no longer had the visibility into like what the business was doing. What I did was I had I ended up working with my like the data analyst team or the business partner to determine like what the right model was to then like, put into production by like so. So there's trade offs. But like, yeah, I would say like learning analysis is like super important, like as data scientists because if you, if you don't understand. Um, for example, like if you're looking at real estate data, if you don't understand that the, the value of the property is sometimes actually correlated anyway with like the zip code and all the other stuff you might see, like, oh, like we got some new like houses into our data set. Look, our average portfolio value went up. Yeah, it went up because you started getting homes listed that were in a very expensive zip code. Right. Your business is not doing better in that regard. You just happen to get. Right. More zip codes. And that's something that if you don't have those analytical tools at your disposal, you might just slap a machine learning model and go like, okay, we're doing great. So I'd say it's it's a super important skill set for sure.

Harpreet: [01:05:24] Akiko, thank you very much. And I hope that I answer your question there. Towson, thanks so much. I'm going to go ahead and start wrapping it up, guys. I've got to head to head to to cause you're going to go see Vir Das live, which would be fun. Got an exciting week up ahead. I'm getting back into recording, so I've got a huge recording spring kicking off, So just want to read through some of the people I'll be interviewing with and hopefully you guys will join me tuned in. I've got to set up like LinkedIn events and I need a virtual assistant is what I need. I'm finding that out because I hate doing these things, like setting up a LinkedIn event that should get to my nerves. So if you are out there and you know a good virtual assistant, [01:06:00] let me know man, because there's so much shit in life that I just don't like doing that. I just, I don't do. And things suffer because of that. I'd rather just be doing the things I like. That's that's freedom right there. That's the real reason I want to get wealthy. So I can only do the things I want to do. Anyways, I digress. So people coming on the show in the next the next couple of weeks, October 3rd, I'll Bellamy October 5th. Megan Lu October 9th. Varun Nair We're going to be talking about his book Breaking Stereotypes, October 10th, the Seattle data guy himself, Benjamin Logan October sorry, that was October 10th. Benjamin Burgundy Sale Date October 12th. Luke Barrows October 16th. Akmal Sayed who's been crushing it on LinkedIn with this content lately. If you guys are not following him, please do. He's got some good. Hot takes spicy takes on machine learning. October 23rd I've got Jessica Ayodele coming on the show as well.

Harpreet: [01:06:53] I'm not proud of the the imbalance in gender on my schedule coming up. I'm well aware of that. I'll do my best to to, you know, make sure we got representation from from all cultures, small, you know, genders and ethnicities and all that stuff. It's just, you know, the people I reach out to don't don't respond back to me sometimes. So if, you know, you know, I'd love to get more more people of color, color, women of color onto the show. So if you know any who are open to being part of the podcast, please let me know, because my mom is a woman of color, so is my wife and is my sister. And, you know, so there's my soon to be niece. So I'm all about supporting and uplifting them. So if you know anyone who who might be interested in coming on the show or you've heard them on another podcast, just set us up with with a quick intro man, you know, like to correct this, this imbalance they got going on, well aware of it and do my best to to fix it. That's it I've got a massive tooth that y'all I've got to redo a root canal. That does not sound fun. My tooth has been hurting for four days straight. Went to dentist today and they're like, We got two options here. You could either pull out the entire tooth. And [01:08:00] I was like, That's stupid because I need to chew or do the root canal again. So I have to do the canal again, unfortunately. But that is it. My friends. Thanks so much for hanging out. Appreciate your spending your time with me and my friends. You got one life on this planet. Why not try to do something big? Cheers, everyone.