>> Hi, this is Jeremy Howard, and you're listening to Coffee Time Data Science, a podcast for data science enthusiasts where I interview practitioners, researchers, and Kagglers about their journey experience and talk all things data science. Before we begin, I apologize for the change to our schedule. Of course, usually you would be seeing Chai Time Data Science on this channel with Sanyam Bhutani.

Unfortunately, he's not available today. He had a prior appointment on another podcast and he was not able to join Chai Time Data Science. So we hope you enjoy this special episode of Coffee Time and Data Science. Without further ado, I would like to invite our very special VIP guest, newly anointed Kaggle Grandmaster, Sanyam Bhutani.

Sanyam, welcome to Coffee Time Data Science. >> Thank you, Jeremy. Usually, I'm very anti-coffee, but I'll have to allow that. I still can't believe you weren't kidding. I mentioned in our message also, I think I don't deserve this, but thank you. Thank you for doing this. Of course, I couldn't say no.

>> Oh, it's my great pleasure. Thank you for agreeing to be an inaugural guest on our show, and apologies for the inappropriate choice of beverage, but this chalier coffee is our national drink, so. >> Actually, at Voke, we were having this funny thread where people were pulling my leg, when you shared the expresso to eat, and everyone was like, "Hey, see Jeremy's drinking coffee.

You should switch over now." >> I will say though, I just purchased three different styles of oolong tea from China. So I also like my coffee. It's very hard to find good tea here though, honestly. So I actually had to get it all specially imported. So I would certainly love to hear about your tips on Indian tea, because I'm much more familiar with the Chinese variety.

>> My mom makes it for me. So every time I have a podcast, she makes it for me even today. She woke up before me. She made two chai, one for prep, one for now, so. >> Great. So that's great for our listeners. If you want a nice cup of chai, head over to Sanyam Bhutani's mom's house, and grab yourself a nice cup of tea there.

Terrific. Well, let's talk about data science. As we said, I guess the stimulus for doing this interview is your recent Kaggle Grandmaster anointment. So why don't we start there? What are you a grandmaster of, and how did you achieve that lofty status? >> So as you can see, I'm talking a lot, and I've become a grandmaster in the discussions here.

So I'm only right now good at talking. Kaggle's belief from left to right is the hardest to easiest category. Jeremy was the first, I think Kaggle Grandmaster, the first ranked in competition. So I'm trying to move towards the left end slowly and slowly. >> I would say datasets contributor, maybe not so hard.

There are probably ways you could game that, I reckon. For example, I bet if I put all the datasets from the course on Kaggle, I would become a datasets grandmaster. I'm one gold away from being a Kernels Grandmaster, which I should definitely get back to doing Kernels, because they're so fun.

So when I ran the Masters Machine Learning course at USF, I actually, the marking for the course was based on how many points you got on Kaggle competitions, Kaggle discussions, and Kaggle Kernels. I always said to them, if you can write in a way that people find compelling, explain data science in a way that people find useful, or build bottles that are highly predictive, then you've got a good opportunities ahead of you.

>> I think sometimes kernels are a bit flaky. I've been spoiled with this box, and I'm so used to things running instantly. Sometimes I get a bit annoyed, but people sharing stuff on there is awesome, and I always learn so much. Every time I go on the kernel stuff, the Kaggle community makes it super awesome.

>> What's the flaky bit? >> I'm just used to running cells super fast, like Kaggle kernel takes longer to start. >> In the actual compute environment, I mean, I treat it as a platform for communicating to people rather than a compute environment. >> For sure. What are you up to when you're not being useful on Kaggle discussions?

Where are you working now? >> I work at Meets and Biases, and we have an awesome fast A culture internally as well. It's like working with friends, working on site projects with friends. I lead our community efforts, which basically means I just do what I love. I host live stream for the broader community, our CEO, Lucas B World, whom you know and our Head of Growth, Lavanya, whose team I work on.

Both of them are super supportive of community stuff, so they just let me do whatever I want. Every time I go with a stupid idea up to Lavanya, taking a stupid idea to her, she's like, "This sounds good to me. If it sounds good to you, just go ahead with this." So I get to do what I love, and for some reason, they pay me for it.

>> Yeah. Well, Lucas is a super awesome human being and also a very smart guy. I know a lot of the fast AI community have made their way over to Weights and Biases and hopefully are doing positive things in that company, which is already going very well. I know it's actually in terms of all of the logging frameworks and deep learning, it's the best integrated with the fast AI library and the most loved by some degree, which I guess is why it's the best integrated, so actually genuinely a good product.

Now, I first came across you when, I think when you posted a message in the fast AI forums a few years ago, introducing yourself and saying you're interested in learning about deep learning. Now, remind me, when would that have been? What year? >> I'm so sorry. Just to add one small point to your previous question.

When I was interviewing for Weights and Biases, I showed this to Lucas and I was going ahead and explaining why I spent so much time on our forums and he just, he stopped me there. He was super appreciative of this, so he instantly recognized that I'm from fast AI.

Now, I get to work on the best experiment tracking tools team. But sorry to answer your question, I got the opportunity to join the International Fellowship in 2017. I was really struggling with the university. I mean, I was doing okay with my grades. >> So what were you studying at university?

>> This is one of my most controversial messages I put out. I was studying computer science, but I wasn't becoming a better programmer, which is how I envisioned it to be. >> At what university was that? >> That was SRM University, so it's one of the good-known universities in India.

I went there expecting to just become better at programming, and I just didn't like the syllabus. I was trying every single thing. I was signing up for every student club and I just couldn't enjoy it. Then I shifted to online courses, was trying to find something interesting there, and somehow magically I landed on fast AI, although initially I was terrified of it.

But for some reason I decided to- >> So let me just dig in a little bit more, because I'm very interested, because I think a lot of people watching will be people who, like a lot of data scientists or people doing data science don't have a computer science background.

I think a lot of them, I certainly used to be like this, looks over at the computer science world and thinks like, "Oh, I probably should have done a computer science degree. Maybe I should go back to university. I'm never going to understand anything. I'm not a real computer scientist." So you're saying that's not necessary, and in fact you're saying you didn't learn much useful programming.

So tell us more. What were you doing in the computer science course? What were you learning and in what way was it not satisfying? >> One of the things I learned afterwards was any computer science degree is more of double ECS, so electrical engineering and computer science. Half of it is just learning about diodes, stuff like that, and that was absolutely terrifying to me.

I had no interest in it at all. I just barely passed in that subject. I remember going to the professor and just asking them, "Hey, please don't fail me in this. I don't want to take this class again." >> Does that work? Is he like, "Okay, I'll pass you." >> Worked for a few subjects for me.

But I looked at the programming courses. First-year syllabus in India is usually just the basics of everything. So they teach you chemistry, they didn't even teach you biology and computer science, and I just didn't like it. I decided to go to the library. I would pick out the senior yearbooks.

I would hang out in the section where all the sophomores were hanging out, and they were doing teenage stuff there. I was the only one with a book in the library in a corner, and even that was very outdated syllabus. So it was just very mundane stuff, and I couldn't see the stuff connecting to real world.

So I would talk to people on the Internet, they were building amazing stuff, and I just didn't see the connect happening at all. >> Yeah. I mean, I remember, so I ended up majoring in philosophy at university, but I did try many things along the way to try and find something that seemed more interesting, and I did do a computer science subject.

It was like a programming and statistics class, and that programming was in Pascal, and I didn't understand any of it, which I found surprising because I knew how to code. I remember going, I found that textbook 10 years later, and I went back and I reread it, and I was like, even though I know it, I've written production software in Pascal at that point in Delphi, I still didn't understand the book, and I realized like, okay, the problem is not me.

The problem is the book. A lot of academics just aren't good teachers. They make things just so hard to understand. >> That realization came so late to me, like two years into the degree, but still like that was super late for me. I was just expecting my teachers to be super good at everything, and I would like go to them with questions.

Sometimes I just wouldn't get answers, and that would really annoy me. Like how do they not know this, like they're supposed to know this stuff. I remember in like one of my machine learning courses, there was a question in an exam about what's the latest here, and I had read a paper.

I drew that architecture. I was super proud about that, and I got a zero because they said, "This is not in the syllabus. You're not supposed to write this." >> Oh, no, yeah. Oh my God. I mean, it's not just university, honestly. My daughter is six, and so she's doing primary school, and it's same issues.

They're like, "You're only allowed to be covering this part." We're talking about fractions, but you're allowed to talk about halves, and if you mention quarters, you're going to get told to stay in your lane. It's bonkers. You were looking for other material, for other things that might help you learn more pragmatic, how to code, how to become a good programmer, and so it was during that search that you came across fast.ai?

>> I first found machine learning, and then I tried. I used to proudly say to my peers, because it was like a status symbol to me, just a turning group, your effect rate. I was at the initial bit, and I would just tell them proudly, "Hey, I did like 40 online courses, and then I would go ahead and sign for five more." >> So it's just like Oceara, or Udemy, or stuff like that?

>> Every single thing, you name it, I probably would have at least watched five minutes of it, and then they did nothing. I wrote an article how not to do fast.ai, where I would say, I would discover something, find a course, study the course, and not be able to build something, and during the search somewhere in between, I found fast.ai, and I signed up for fast.ai, luckily.

>> What was that experience like? Was that different to other courses you're taking? >> I could feel like this was the first time I felt that I could build stuff that actually works, so it's not just those three layered neural networks. I remember in the first two lectures, you showed us how to get started on a Kaggle competition.

We put out a simple baseline, and we were beating you. So I remember me and my roommate were taking the course together, and we both were jumping for 10 minutes. "Hey, Jeremy, how is this happening?" on the leaderboard. So here is the aforementioned hownottodofast.ai, which I do remember. It's very, very helpful, and I've shared this with other people, and I think Radek has also shared some of these ideas.

It's interesting to me, yeah, it's interesting like how few, even machine learning courses are top-down. So I always feel like if you're not creating useful models reasonably quickly, then that's pretty discouraging. And also, how are you going to know what's the point of the stuff you're learning about if you're not actually training useful models?

It seems like it's pretty hard to integrate the knowledge that you're being taught. >> All of the courses would-- Sorry. >> Go on. >> All of the courses would like, now most of them are better, but at that time they would start with showing this super cool style transfer example, and then they would go back to just teaching the mundane stuff, how you do stuff in Nampai.

I was so done with that at that time, and Fast.ai was super awesome because there was also this insane community, and usually when you take a course, you're in a similar cohort, which is one thing I didn't like about university. No one around me was talking about machine learning.

They thought I'm super weird. I thought they're super weird. But on Fast.ai, we had-- >> What were they wanting to do? If they're not interested in ML, what were their hopes and plans and interests? >> Just do T-Net stuff, as I mentioned. >> So it wasn't just ML, but it's more just actually studying and learning effectively in general.

>> Yeah, and most of them were just building websites. I didn't enjoy that bit at all. I would just download templates and hack stuff together every time I wanted to do that. >> Yeah, building websites is not particularly intellectually interesting of itself necessarily for sure. So yeah, tell me about, okay, so you got started and you said 2017, right?

So how did that go from there? Was that then all smooth sailing, or did you hit any obstacles along the way, like anything that in hindsight, you wish you had done differently? >> So you had already pointed to my article, and I also was asking this question when I had the opportunity to interview Rachel, because I'm so new to top-down learning still.

In my entire student life, which is 15 years I've been studying in the bottom-up way, and I was so new to top-down, I would always default to that. So my issue was not listening to you enough. The success I have was because I listened to you 10% of the time, and other time, I would just like, okay, this is not working, I need to go and read the basics, or again, default to bottom-up learning.

>> Exactly like Radek Khosmalski. He says the same thing, yeah. >> We both also talk a lot about this. He even, we spoke about this for half an hour when I had the chance to interview him. Even in this book, I think this is really a struggle, because we always start with aspiring to be someone, and then we don't want to put in a lot of effort initially, and you don't see the dot connecting immediately, although it takes an insanely long amount of time.

>> So can I ask, and feel free not to answer if this is too private, but how old were you in 2017 when you joined the Fast AI community? >> So five years ago, I was 19. >> 19, okay. 'Cause I ask because I, you definitely stood out to me even then, you know?

Like, and I couldn't exactly tell you why or how, but you certainly came across as, you know, unusually motivated. Now I will say, that's not that unusual. There's plenty of people who joined the course sounding pretty motivated, but then I would say like the big difference with you, the really big difference with you is your tenacity.

Like you didn't go away. Like, you know, like if the multiple 100,000 people who start the course, you know, seeing it through over a multiple year period and doing the things, like I remember you would even post on the forums and say like, you know, in order to have something to force myself to live up to, I'm going to tell you what I'm going to do over the next few months.

Like, there it is. What, do you have a sense of like, what's, how did you get to this point? Why are you more tenacious than the average bear? And, you know, what's kind of helped you to stick with it and to follow through on your goals? - Yeah, you're being too nice to me.

So like, I would go away from Fast Day because I was like, maybe only Jeremy can do this stuff. I'm not like as good as you of a programmer. I'm not as good as you at reading papers. So I would like go away to other courses time to time.

And again, like realize, no, I want to come back to the community and just the mad passion of really believing that if I stick to this would be really helpful. One thing I've learned is like, we always start aspiring to be some person as I was saying. And we're like always looking at that end goal, but instead I've over time learned to really enjoy the process more.

So like when I was trying to lose weight, I was like, that was my fifth attempt at losing weight. This time what made it stick was I would just show up every day and walk for like two hours straight. Like no excuses, just walk for two hours for at least four months.

And then I started losing weight. So broadly speaking, I just learned to appreciate the process of showing up every day and just doing it. - Yeah, okay. That's great advice, is to focus on the process. I had a similar experience, gosh, how long ago? Seven or eight years ago, when I decided to try to improve my fitness and I started going to CrossFit, like CrossFit was literally a block away.

And I had really never done any kind of intense training in my life. And CrossFit, you dive very quickly into intense training. And each day I just felt awful. I can't begin to tell you how much I hated it and how bad I felt. And the other thing that makes it difficult was like everybody else at CrossFit is just so like gung ho about it.

So I think I must've felt a bit like you was like, oh, Jeremy's better at coding than me. I can't do it. It's like, okay, these people are all CrossFit people. But yeah, you know, it helped to have Rachel actually 'cause we both decided to go every day. So like, okay, we're gonna go every day.

I think it was like 5.30 AM, turn up, hate it, go home, turn up, hate it, go home. And after about similar time, I reckon about three months, I was like, I still hated it, but I didn't feel horribly sick every time. You know, I was just like uncomfortable and a little painful, you know?

Yeah, it's hard, isn't it? Because like to get stuff out of these, any of these things, you have to turn up to the process for months and those months are often not much fun. - CrossFit is also super hard. You take, in one day you get super exhausted and next day you have to show up again, half exhausted, you take thinking results aren't coming in.

What do I do? - Yeah, and I mean, so like when we're learning stuff, I think in university, it often seems to be the professors are proud of this, you know, like weed out classes. They're like, oh, 80% of my class flunked out. It's like, oh, you're a shit teacher, aren't you?

You know, like you wanna be giving people positive, real positive feedback all the time of like, oh, wow, you just trained a model. Oh, wow, you just fine-tuned a model. Oh, wow, you just created a web application, right? That's how computer games work. You know, with computer games, you don't like start up Gran Turismo 7 and it's like dumps you straight into an F1 car on Nürburgring and asks you to, you know, qualify for, you know, F1.

It's like, okay, we'll start with license test and I'm not gonna afford focus. Can you drive in a straight line for 50 yards or something? And at the end it's like, congratulations, gold medal, you've done it. Okay, now we're gonna like show you how to do it and you have to press the brake pedal yourself.

It's like, congratulations, you did it. - I think one of the things I learned through gaming, I used to game a lot. I would like sit down and game straight for 16 hours. My parents weren't happy about it, but like I would just fight and I would fight it, make that happen.

In games, we call it grinding, where you like have to do this so stupid stuff over and over again, that just doesn't make sense. But you enjoyed so much and like just, I think many Kagglers see this, that like bringing that over to Kaggle, because initially you just suck on the leaderboard.

I still suck on the leaderboard, I'm like bad competitor. But you enjoy the process so much because it's the grind that you start enjoying towards the. - Now you say that, but I don't think that's true 'cause you've had some very good results in competitions, haven't you? - I had the chance to team up with awesome people.

So I was learning from them and contributing ideas. (laughing) - Now, the other thing I noticed with you is in terms of follow through, is you know, we often encourage people to kind of put themselves out there, build their personal portfolio, blog, whatever. Everybody says they're going to, not many people do.

And it's hard, I know like you've talked about how, for example, this podcast, when you started, no one's listening, you know? It's just you talking out to the empty internet. Like give me a sense of like, what did that feel like to like, you know, write blog posts when you didn't have a readership, create podcasts when you didn't have listeners?

Like, what did that feel like? And how did you push through? And then how did you get people to start noticing you? - I'll start with the tangent. I think great people improve and like they reach heights. Greater people keep, they uplift others and greatest people keep reaching higher heights in life and keep uplifting others.

So people like you and the Kaggle community and Fast hair community are the greatest because you keep uplifting others also. My blog got a lot of traction in the podcast because you and the community were sharing it so much. But it was like super bad for me. As you might know, I'm like extremely sincere towards anything I try and like I give it my best.

Otherwise I'd like just wouldn't sign up for anything. That's how it is now initially it wasn't the case. - Let's see, tell us more about that. Tell us more about that, what do you mean? So, I mean, that's just such an interesting point. So you're saying nowadays when you try, you really try, but that wasn't always the case.

Tell me more. - I would just like bail halfway through like three days into it. One month into anything, I would just bail. When I was trying to lose weight in university, I would like work out. I worked out for one month super intensely. So I decided I would climb, we had super tall buildings.

So like 15, 16 floors. And I used to live on the ninth floor of the hostel. Out of this like nowhere, I decided I'll climb 60 floors every day, at least 50 to 60 floors. I did that for one month, damaged my knees a bit. And then I just like gave up.

I didn't like start slow and build up. - Right, so what happened, you just became more mature or was there some kind of like conscious decision or? - I think maturity, yes. I'm still not very mature, but now I like start slowly and like remain at anything I do.

- Okay, so sorry, so go on. So you're very, so at the point you started the podcast, you were sincere about putting your effort in and yeah, go on. - Yeah, so when like I got to interview you or other agri-grandmasters to me, it felt like I'm being insulted towards people like you because it's not getting too large enough audience.

And I felt like that's my job to like, if I'm asking for someone's time, I need to like put it out there. And that part was super annoying to me. Like how can I, this is something I'm putting out to the best of my ability. How can I like not get it to the audience that I think it deserves?

- Right, so how did you? And yeah, I mean, like even just to put out that next podcast or that next blog post, yeah, how did you convince yourself to kind of do it anyway and then did that audience like just gradually trickle in more and more or was there something that was like, suddenly you got noticed or yeah, what happened?

- I learned this from Radeko Smulski. He would post his goals and that's what I started copying. So from him, I started, as you mentioned, I now post my goals every single year also. I just, now I also make a video about it so that like everyone, I'm terrified then.

Like I think everyone's watched that video. Now I need to like do that. And I switched from thinking about what's the outcome to like, I'm going to, when I was doing the podcast, I decided I'll put out two episodes every single week, no matter what for like an entire year.

I just decided to do that. And I wasn't so much so looking at how many people are watching, although I would like pay some attention to it. And instead I would just like focus on making the best version of the podcast every single week. Or before that I was writing blog posts and I committed to writing one blog post at least every week.

And I was just making sure I can write anything best to my ability at the time. So here is Radex's book where, yeah. Highly recommended for, he's had a lot of success and has a lot of great tips. I thought I'd better just share that since we've both been talking about Radex a little bit.

Yeah, it's what you described, you know, which is like just trying to do a better job, you know, to get there. It requires a certain amount of kind of like trust in the system that there is some reasonable relationship between working hard and getting good results, you know, which I think is not obvious, you know, like it's not obvious that that's true.

And so, and if you don't believe that's true, then it's gonna be very rough, you know, because it's just gonna feel like, well, why bother? I guess, you know, actually I wanna share something which is I, to give a sense that it is true. And so I mentioned earlier that I'm close to being a Kaggle Colonel's Grandmaster.

Here is the last few Kaggle Kernels I've uploaded. And you can see every single one of them has got a gold medal, which is to say like I put my all into creating the very best Kaggle kernel I could. And yeah, it's like, and each time I have done that, people have reacted to that by upvoting and liking my Kaggle kernel, you know.

And so like, I think like, if that doesn't happen, it's important to self-reflect, you know, same in a competition, a Kaggle competition. Like I like these transparent metrics 'cause you can't bullshit yourself about them. Like, I mean, you can, but you shouldn't. It's like, okay, if you don't get likes on your Kaggle kernel, you should assume 'cause it's 'cause you didn't do a good enough job.

You know, you didn't explain it well enough or you didn't make it compelling enough or it wasn't a particularly good approach. You know, if you sucked in the private leaderboard, you shouldn't assume you just were unlucky, but that, you know, this is an opportunity for you to learn how to do better.

- So sorry to disagree a bit, but initially, there's like always this period of suck where things wouldn't just pick up, at least on Kaggle also, when you're just putting out stuff or kernels, I think it takes a while for people to start voting it. So I would also like get-- - Well, that's true, that you have to notice you.

Not in competitions, you know, not in competitions. Yeah, I mean, certainly in podcasts and blogs and things, it's very different. It's like, trickles up very, very slowly. Like people often ask me how to get or Twitter followers. And I'm just like, just tweet. Like just, you know, and to try to make good ones.

Like I started with zero followers and then one and then two, you know. It always goes very, very slowly. - I'll share this one thing from my, so this is titled as My First Kaggle Competition Experience. And on the leaderboard-- - How much is your name? (laughing) - There's a story behind it.

We were so naive. We were trying to rename stuff and my teammate ended up cleaning up my entire laptop without command. (laughing) This is how the first competition felt. Like I was looking at the Kaggle Grandmasters and it was like, they're on a supercar, I'm running barefoot and it's a 100 miles sprint.

I can't keep up with them. I would submit to the leaderboard, sleep. And as I wake up, I would have fallen down 50 positions. I would come back, try it again, keep doing that every day. And like, I finished in the top 20% I think, which also was like super awesome to me.

- Fantastic, and again, that's like tenacity, right? Like most people aren't gonna keep coming back. And in my experience, the people who do come back, make it, you know, like it's, I feel like this is the number one difference between those who succeed and those who don't. But yeah, coming back to your point about like for example, getting people to upvote your kernels or to watch your podcasts.

I guess what we're saying is having extremely good content is a necessary but not sufficient condition for success. So you do need to make sure that you're telling people about your Kaggle kernels or about your blog posts or about your podcast episodes. But if you get people to look and they're not extremely good, they're not gonna come back and look again.

- One thing, and to credit, the blog started through Rachel's advice. I would just read it every week at least once and I would sincerely follow her advice on how to get started on blogging. It's an amazing blog post that she's written. That's how it got started there. And I would just transfer my learnings to the podcast also.

But the podcast was like super bad in quality during the initial weeks, but at the time that was the best I could do. And one thing I've learned is I try not to spam too much with like sharing it over and over again. So every time I've done a podcast, I've only shared it once with the world.

And I, if I re-share it, it's because it's so relevant. I say, "Hey, please listen to this. "This is where I discussed this." But I'm also like extremely sincere about not spamming the community. Although like I do spam with my tea puns and tea jokes just for the sake of branding, but I try not to spam too much with like any stuff I'm building.

- So this is Rachel's blog post, which I've heard what you just said. I've heard it from so many people that this is what got them blogging. And you know what a lot of folks might not realize is Rachel, you know, she's a math PhD. She's an academic. The idea of putting yourself out there like that is an absolute anathema to her academic training, you know?

And, you know, in the academic community, which I'm also in nowadays. Yeah, it's almost look, it is. It's looked down upon, you know, of like communicating in a clear way that the rest of the world can understand, which is, it's such a shame, you know? And so it does require a pretty conscious decision, I think, particularly if you're at university to be prepared to be different.

Because the people teaching you and supervising you have probably never written a blog post in their life. And they probably have no idea what it really is. And they're trying to get you to write the most obtuse academic jargony prose in extremely exclusive PDF-only academic papers. So tell me more about like creating educational content in general.

What's the kind of educational content that you're finding is really gelling with people, that people are telling you, thank you so much for creating that, that made a difference to me. I'm still surprised when people say that to me. I'm like, I'm just creating this for my own self.

I just follow Rachel's advice, which is to create something that I wish was there. And that's how I started the podcast also. I was talking to so many incredible people on Fast Day. To Atini, who used to take the course, he taught me how to get started in freelancing.

And I would always shamelessly ask question. I would always start with an apology. Hey, sorry, this is going to be a super stupid question, but I want to ask this to you. And I would like just shamelessly approach people, ask these questions. That's how the podcast also got started.

So I'm just trying to fill the gaps that I think exist. We have already. - It's something I learned. Something I learned in my 20s is, yeah, that which you've obviously already discovered, is that those embarrassingly stupid questions are the questions that lots and lots of people want to ask, but they assume everybody already knows.

And so they end up never getting asked and everybody's like, we really don't know. So, speaking personally, when somebody asks me those questions, I'm always very grateful. Because I'm always like, ah, that's a good point. I had, you know, like, particularly if you're an expert on something, you don't really think to mention it until somebody asks you the question.

- Yeah, I've been super blessed that people have been so nice and they always share their knowledge like they've shared it privately with me. And also like now on the podcast, they've been sharing it. So I'm super lucky in that way. - Have there been things that have surprised you either that have been very popular that you didn't expect or some things that you thought like, oh, this is gonna be big and they didn't really go anywhere?

- Honestly, I started with the arrogance of like 20 year old, hey, people will listen to this. I know this is awesome. Everyone should listen to this. This will be the number one data science podcast in next year. No one listened to it. And then I realized as I got a bit more mature that it's super hard to get people to click on and our long video on YouTube alongside with everything else that exists on there, especially stuff that's like super technical, super knowledge heavy.

Now I feel super grateful that people, thousands of people listen to the podcast. They click download, they click on YouTube. So now I feel like. - Yeah, and I think people are like, I think it's under appreciated how much people do want that kind of content. Like, you know, you look at something like Lex Friedman's podcast and it's not flashy, you know, like it's well produced in that Lex spends time on like, he has high quality equipment and he sets it up carefully and he does it in person.

But you know, his actual interviews is just like him asking questions, generally fairly brief questions to somebody who's generally highly technical listening to the whole answer, rinse and repeat for one and a half hours. And it's, you know, lots of people do actually end up tuning in because they're, you know, not everybody, but there are people out there who want high quality, not dumbed down content.

- I remember watching your interview on his podcast and I thought, no, you'll say yes to me. And when I said I want to fill the missing gaps, that's what I meant. Whenever I do a podcast, for your case, I listen to every single interview, every single email you've done.

And then I would ask the questions that weren't discussed there, so just try to bring those things out and not always start with the repetitive questions that many people do. - That's a great point because Lex does that too, you know. And I don't do many interviews at all.

Like I did one with Lucas, you know, who runs Wets and Biases, I did one with Lex. I mean, hardly anything else. Like, and in both cases, Lex and Lucas like, just had done their homework so well. And so the questions they asked were genuinely interesting. And so to the kinds of people who are interested in my thoughts in Korea, it's gonna be interesting to them because it's just, it's not just like, oh, who are you?

What do you do? It's like, oh, do this thing called fast AI. And, you know, anybody who knows me doesn't wanna hear more about just like, what's fast AI? They know what's fast AI. So, you know, what you're describing, it's how everybody should do it, but hardly anybody does.

(laughs) And, you know, probably a lot better for you as well, right? 'Cause in the process of doing that research, you're gonna learn about stuff. - I personally feel maybe it's because of my culture and upbringing, it's also insultive to the other person to ask like very basic questions.

Like if I interview you and ask, hey, Jeremy, can you please introduce yourself? I won't do that. That's my job. I should be asking the interesting questions to you and to anyone that I-- - I think it's insulting in our culture too, but people still do it. (laughs) Anyway, good on you.

I mean, something I do wanna ask about, which I definitely remember is the excitement that you shared about having the opportunity to interview for a residency at Google. And I think, you know, the community was excited for you as well, 'cause it's like, wow, you know, good stuff, Sanyam, this is gonna be great.

And then it didn't happen. You know, tell us a bit about that. Like, what did it feel like? Like, how did you get that offer, you know, to interview for Google? What did you do to prepare, you know, what did it feel like to fail at that, you know, at that journey?

And how did you, you know, push on from there? And what did you do with that experience? - I was crying at 2 a.m. when I got the email that I had rejected. - Yeah. - And I used the word fail very intentionally, right? 'Cause that's what it is, and that's what it feels like.

And it's like, okay, I've worked this, I failed, you know? And that's like, you know, let's not use bullshit words about whatever it's like when we fail. How do you, yeah, how does it feel? How do you get up? So you, I bet you were crying middle of the night.

Probably felt like that's it, right? You tried so hard to get here, it didn't work. You're not good enough. Is that kind of the feeling you're having in your head? - Yeah, and the point I was getting to was also that I was again so distant to my peers.

I remember telling my friends, hey, I interviewed at Google AI residency and they feel me. And they just say, cool, okay, you'll get it next time. Like they couldn't appreciate how big of a thing it was. - Yeah, that's hard, yeah. - And that's why I'm really grateful to FASTA community when I shared that, hey, I'd failed and I've done bad for the community because I couldn't get through.

Everyone was like super supportive. That instantly got me to getting back. And in fact, that's how I started the podcast. I decided, okay, now I'm going to help the community to at least get to the point of interviewing and hopefully someone gets through. - Fantastic, so, you know, that's, I mean, I'm sure there's a lot of genuinely important outcomes, but how did you get there in the first place?

You know, like you say, you come from a community where nobody even knows what you're doing, let alone gives a shit about it. How did you find yourself flying to America, to Google? - I just, I found this courage through applying through FASTA, I didn't expect to get into the course even.

And since then, I've just found the courage constantly to seamlessly apply, seamlessly ask for help. So just stepping out of my comfort zone. I'm very inherently shy, I've been shy all my life, but now on the internet, I'm less shy and I just apply to every position. Even in my undergrad, I would start applying to positions that I knew were clearly out of my league.

And my goal was to just get to the interviewer and ask them, hey, how do I like actually get through this interview? And I never got a clear response at that time, but I just got to the extent where I could comfortably apply to things I didn't expect to go through.

- And every time you do that, you know, there's a chance you'll fail, right? And so, but if you don't apply, then you always fail. So, yeah, I mean, I'm the same. I am introverted and shy. And I remember telling my friend, Chris Latner, that a few months ago, and he was just like, sorry?

I was like, I am. And he's like, that can't be true. He was like, here's all the things you do. I was like, I know, and they're all terrifying. And after it, I'm exhausted and I do it anyway. Yeah, you know, you have to put do it anyway, don't you?

You know, 'cause otherwise you'll always fail because you never give yourself the chance to succeed. - Yeah, and just staying at it, I think I'd agreeing to the fact that initially I, of course, at that time didn't believe that I can't get a job just because I was so arrogant in my head.

Here, I'm ahead of my classmates. Of course, I should get a job. India is a big country. It's very populated. There's a huge amount of competition here. So I, of course, wouldn't get the jobs. And I would continue in my arrogance somewhat in my tenacity to continue talking to people, continue applying left and right.

- Yeah, so, okay, so you started the podcast and, you know, what happened? What was the next commercial opportunity you got and how did you make that happen? - I started it inspired by you where I decided not to monetize it, which wasn't the best decision because I was investing my money and not getting any returns.

My parents were definitely upset. - Sorry. (laughing) - No, but I didn't mean it in that way. So initially, I started the podcast and then someone at H2O, H2O AI, it's an awesome company, shared my podcast internally. They were like, this guy's interviewing everyone. I want to interview people at our company, but he's interviewing them before I do.

That made the CEO reach out to me. The CEO of H2O decided I want to hire this guy. And then I got it to do as part of my day job. - And he's another Indian guy, right? If I remember correctly. - Yes. - Yeah, cool. - Instantly, that's how the dots connected.

And through any job I have had, this is I think the fourth stint in my career. It's always been through the FASTA community directly or indirectly, I feel. - Yeah, no, I mean, that's actually amazingly kind of direct is like your personal branding. I mean, not that unusual though, you know, your personal branding efforts was the thing that made people not only know who you are, but you know, that you're doing something that's so good, like that's literally what they were wanting to do themselves.

That's amazing. Okay, I have reached the end of my questions. Did you have anything either that we haven't covered you'd like to cover, or is there anything that you would like to ask me before we wrap up the inaugural coffee time data science episode? - I would ask you for advice.

What do you, as my guru and my teacher, what do you expect me to do next? What do you want me to do next that'll help make you happy? - Oh no, I never have opinions, you know? I try to like, I try to kind of say like, here's how you can move in a direction that you want to head, you know?

Like, I don't have any sense of like what direction should anybody head. I mean, hopefully, in a direction that's at least not destructive to society, and that you've got the tools and thought processes in place to think about how to help society. But like, you're, I mean, you're doing exactly what I would have thought makes perfect sense, which is to surround yourself with not just good people, but people who appreciate you and value you for who you are and what you do.

You know, and particularly, you know, in a geography where that isn't gonna happen just by virtue of the people who happen to live down the street, you know? You know, the only thing I kind of tell people of around your age is you could consider changing your geography at least for a year or two.

Like, for me personally, one of the few regrets I have in my life is that I didn't do that. You know, growing up in Melbourne, very different to India, but in some ways, you know, it's got some similarities in that it's a long way away from, you know, at least kind of culturally and in Australia's case geographically from anything like the kind of stuff that I was interested in.

And nobody in my life cared about anything that I was interested in. And that made me feel like there was something wrong with me. You know, like I was weird and that was a problem. And, you know, people would sometimes fairly directly give me that feedback, you know? - That sounds very similar to my experience also.

- Yeah, so going to San Francisco to me, like there's a lot of things I don't love about the US in general in San Francisco in particular, but it was such a confidence building exercise to be surrounded by people who were interested in what I did and did try to do similar things themselves.

And to also realize that the people that I admired and looked up to were not on some other level of existence. They went on some other plane, but they would, you know, they made mistakes and had setbacks and redoubled their efforts and just normal people, you know? And so I kind of thought like, okay, I'm, yeah, they're just other people like me and I can do interesting things like they're doing.

And like I'd always wanted to do interesting things like they were doing, I guess. So that'd be the one thing like to consider. And I'd say like a couple of years in a invigorating culture full of people that respect the kind of work you do can be of value.

It's not possible for everybody, you know? And honestly, it wasn't really possible for me for most of my life. So it's not really a regret in that it wasn't something that I had access to or something I wish I had access to. And it's definitely not needed. Like you can totally do it all, as you've shown online.

But yeah, it's something to consider if it's an option. - Thank you. Now I've gotten to the stage where it's somewhat possible. I'll try to explore that option. But before we wrap up, I also want to really thank you for all the knowledge you've been sharing with the community.

Your interview will always be pinned on my channel. This is the 150th episode on the podcast, but I'll always be grateful to you and the community. All of my small achievements are through Fast Day, so thank you so much for- - It's actually the first episode of Coffee Time Data Science, but I know what you mean.

Well, we'll let you keep it as the 150th. You're more than welcome. And thank you, because I mean, you've given back so much to the community. So I know so many people are grateful to you. So thank you. - They've been too kind, just to add one more thing.

I know I have a lot to learn still. I've gone off on this tangent of creating content, and I feel like an imposter that is only a discussion's grandmaster, is only a competition expert, and that knowledge needs to transfer somehow. So I know I have a lot to learn still, and the imposter syndrome is taking over, but I'll- - Wow, you're still young.

You'll get there. Thank you, mate. Have a great day. - Thank you so much, David. - Bye.