Andrew: Hey there freedom fighters My name is Andrew Warner I’m the founder of mixer G where I interview entrepreneurs about how they built their businesses and a lot of times when I talk to These entrepreneurs they talk about how they use data because we’re all so cerebral.
We’re thoughtful. We’re intentional and data just makes a lot of sense or At least it seems to when you hear somebody in an interview and then you go and you look at your own data You even look at basic data like what google analytics and you go on here And you kick it off to somebody else who could figure it out or you think all right?
I’ll focus on something else like customer experience and you blow past data and that’s a common experience Which is why i’m happy to have for today’s guest amit Prakash, he is the co founder of ThoughtSpot. They, uh, they’re an AI powered analytics company. They help users make the most of their data, and I love the fact that it’s AI because it means that it’s gonna be like talking to a human without actually talking to a human.
Amit, I see you’re smiling as we’re talking about this. You recognize some of these problems. You know why the solution is so appealing right now. And I want to thank the two sponsors are going to make this interview happen, where we can find out how he built this company. The first, if you’re interested in a decentralized autonomous organizations, I’m going to tell you later why you should check out my podcast.
It’s available at joinorigami. com slash podcast. And second, if you’re hiring developers for AI or anything else, go to lemon. io slash Mixergy, but I’ll talk about those later. First, Amit, good to have you here.
Amit: you for having me. Looking forward to the conversation.
Andrew: I’ve seen all kinds of valuations of the business. Let’s give people a sense of how big you are. Can you talk about what the valuation is?
Amit: Yeah, so the last round that we raised was at, uh, 4. 2 billion post, uh, money. And, uh, we’re about 800 people in the company. We’ve been around for about 11 years.
Andrew: and revenue. Can you say what that is annually?
Amit: Uh, it’s somewhere around like 100 million ARR.
Andrew: It’s a phenomenal business. Incredibly big. Let’s talk about what they do or what you do in the context of an example. Do you have an example of someone who’s used ThoughtSpot in a way that we can understand what you do?
Amit: Yeah, so I’ll give you a few examples. Um, I think one of our earliest customers was. probably the world’s biggest, uh, retailer. And, uh, what they were experiencing was that, um, when their team tried to make decisions around merchandising and promotion of products and what to promote on their website and which inventories to clear and things like that.
Um, they were experiencing delays up to a month in getting all the data together. So their team could only meet once a month and make these decisions. And those decisions were locked for, you know, a month. And when they compare themselves to their digital counterparts, these things are happening every hour or so.
And the key problem was that they were using sort of previous generation of data and analytics tools. And so when somebody had a question. about data they needed to file a request where an expert was going to go and build this and they had a long queue and things like that. So decided to go with us.
And the key thing was that people who are in merchandising jobs, whose job it is to figure out what to buy, what to sell, what to promote and things like that, could just, um, in basically a blink of eye, get the information they needed. And so they deployed us and they were actually able to shrink their decision cycles from a month to essentially an hour.
And that was a huge competitive advantage for them. I’ll give you a different example.
Andrew: Before you go to another example, how long ago was this?
Amit: Uh, this was somewhere around 2015, I believe.
Andrew: So, I know this company. We’ve agreed that I wouldn’t reveal who it is. It’s a well known company. My audience sells on their platform and knows them really well. They probably are customers. I thought this was a core competency of… every internet company since at least the year 2000, if not the year 2010. How is it that they, what are they doing, how are they even able to sell without having the ability to adjust based on data?
Amit: Um, so I think at this point, their online business was not the main driver. It was kind of, uh, a side thing, perhaps for the main business and I wouldn’t be able to say anything more than that, except that we knew this was a problem, and that’s why they approached
Andrew: But maybe you can just take me into, I don’t want to look down on them and go, ah, these big companies, they don’t understand anything. I want to understand them instead of pushing them off. What is it like for an organization that big? How do they operate? And, and what works without data? How are they able to post things on their website and to change it?
Was it gut instinct? Was it more like a tastemaker internally? Or how did they operate before ThoughtSpot?
Amit: No, so they were absolutely data driven. It’s just that, you know, um, using a set of tools that were not designed for That scale, somebody had to, somebody else in whose job it was to data had to run database queries, extract that data, fed it into something like a tableau. And then build some sort of a report and there are probably 500 people doing the job of merchandising and maybe 20 people with data titles supporting them.
And so there’s a huge imbalance between, um, sort of the demand and supply. And, uh, both sides were unhappy because one side was not getting the data they needed. The other side was kind of feeling like they were going through this endless, thankless loop of pull this data for me, pull that data for me.
Andrew: So what did ThoughtSpot do for them? Did it just pull reports faster? Or was it even taking over the homepage and displaying products based on data?
Amit: Uh, no, so we had no, uh, visibility outside the company. It was just helping those merchants. Uh, merchandising experts make the decisions much faster. And the way we were able to enable that was first, um, the, our system was scalable and could work with granular data at transaction level. And second, we had this Google like interface that’s fairly easy for most business people to go and ask questions in.
Um, so they were not reliant on this 20, 30 person team. to be able to get the data to them. When they thought of the question, they could go and ask the question and get the answer. And if they looked at the data and that caused another question in their mind, they could immediately go and get answer to that as well.
Andrew: And so, can you give me an example of the type of question they might have gone to ThoughtSpot and, let’s say, almost Googled for an answer in the data?
Amit: Yeah, so typically they would have, um, a set of product categories or subcategories with a list of SKUs, and they would want to see, um, how much of this sold this week. How does it compare to the exact same week last year? And if there are changes, where are those changes coming from? Is this a particular demographic that’s not buying?
Is this a particular geographic location that’s not buying? Or, um, is there a particular product in that category that’s just shooting up dramatically? One fun story that I heard was that, uh, when fidget spinners became a thing, they were able to detect that several weeks before. Otherwise, it would have happened and they were able to.
Kind of stalk it.
Andrew: Uh, I see. You know what? Before we continue with more examples, I mentioned Googling partially because you were a tech lead at Google just before you founded this company. What were you doing there? And then we’ll talk about how you came up with the idea or how your team came up with the idea for ThoughtSpot.
What was it? What was, uh, your work at Google?
Amit: Yeah, so I worked at Google for about five years leading a team that was responsible for predicting how likely it is that somebody is going to click on a particular ad, um, given what we know about them, where they’re looking at the ad and what the ad is. Um, so this was, as an engineer, like an amazing opportunity because we were training machine learning models.
at two orders of magnitude larger scale than anybody else in terms of training examples. Literally like in a year, there’ll be trillion training examples and a really wide feature space with possibly billions of things affecting what could happen with an ad. And it It directly connects to Google’s revenue engine.
So every time… So I was responsible for the AdSense part of the business. So every time my team would be able to improve the quality of these predictions by a percent, it was effectively a percent more in revenue from AdSense. And, uh, Our goal usually every quarter would be to improve it by a couple of percent.
Get sometimes 1%, sometimes 1. 5, but consistently quarter over quarter these improvements added up. So it’s a really fun job.
Andrew: We’re talking about hundreds of millions of dollars in extra revenue through each improvement and the improvement that a human being might make is to say change the color of blue and that was the early days, uh, Google. Software would do something much more intelligent and customized for each individual user based on what it’s learned about the user, about the advertiser, about the spot that it’s on, the time of day, and random things that we can’t even anticipate, right?
Amit: so it would be subtle changes, but it would ever so slightly increase the relevance of the ads to the user. Um, so, so for example, At some point, someone decided that ad blindness was a thing. And so we started measuring if you show the same ad to a person several times, what happens to
the probability of click through rate in terms of how it starts diminishing.
And so the right thing would be that after a few times, give some other ads a chance for that particular user. And, uh, it all just happens through a mathematical optimization. So once you’ve shown this three times, the probability goes down slightly, some other advents in the auction and the diversity starts to kick in.
Um, this whole thing is kind of starts from a hypothesis, then running small experiments, then scaling those experiments to 100% of the traffic. And maybe about. One in twenty of such idea would actually materialize in anything. So, so you basically are constantly running experiment after experiment. And if you run a hundred experiments in a quarter, maybe five of them will give you the game you’re looking for.
Andrew: Did the hypothesis come from your team of, like, five to eight engineers, or is it the software coming up with the hypothesis?
Amit: Uh, no, it was mostly human intuition. Um, sometimes data driven human intuition. Um, and sometimes just… Kind of playing with the product or thinking about how the world works.
Andrew: So it was a great company to work for you’re making huge, uh improvements on an already big business What made you decide to go off and start a new business?
Amit: So, um, early on in my life, my father was a professor in engineering college, and I always wanted to be a professor. And after my PhD, working in industry was supposed to be just a sort of stopgap thing to learn a little bit about interesting problems to work on and research and then go be in academia.
Um, After working at Google, kind of, uh, that perspective changed and I realized that I was working on a lot more interesting and fun problems than what I could be in academia. And so I wanted to work in the industry, but then there was kind of almost like a void that all my life I was pointed in this direction.
What do I do now? And, uh, entrepreneurship kind of naturally came because it’s hard not to be inspired when you’re working in such companies in the valley. So I decided that I was going to do something and the question was what problem and with whom. So I ended up meeting my co founder Ajit sometime early 2012 and we brainstormed a number of ideas.
I was coming from. doing lots of analytics to be able to come up with hypotheses to try. And so data was a big part of my job and Ajit prior to co founding Nutanix, he was head of product for a company called Astrodata and before that at Oracle. And so he had strong background in data. And when we looked at this space, um, we realized that A, uh, both Tableau and Click were doing really well at the time because they were promising sort of unprecedented access to business users that wasn’t there before.
But most of it was still fairly complex and long sort of training classes for business users that they didn’t want to be in. And, um, the scale of data that you could analyze in these tools was fairly limited. So we saw an opening there and, um, there were funny stories about how we, we went to one of the user conferences for these companies and one of, um, their reference customer on the stage was saying that these tools are so amazing that now every time somebody needs a new report, all they have to do is fill up a three page form and within a week the report is available.
And so we chuckled to ourselves and said, We can probably do much better than that, and that was kind of the seed of the idea.
Andrew: Once you had the idea, what’s the first step you took to get it going?
Amit: Um, so here I really have to, um, acknowledge, um, Lightspeed Ventures and, um, Ravi and Arif, who kind of believed in us from the beginning. And, um, We basically, in the beginning, it was just a slide deck and our backgrounds and they decided to bet on us and they’ve been extremely supportive through an entire decade.
Um, and, uh,
Andrew: So you raised money before doing anything else. It was such a big problem. You needed to raise money
Amit: yeah.
Andrew: Okay, and then how long did it take you to build the first version?
Amit: It was, um, so I think, uh, we started in, uh, June of 2012. And we had our first alpha customer in November of next year. So about a year and a half.
Andrew: Okay, how’d you get your first customer?
Amit: Um, it was, uh, one of, uh, Ajit knew them, uh, through some of their connections. And, uh, they just happened to be sort of the right size, right problem. Um, it was, uh, It was the marketing team of an accounting firm, and they were running their business on MS Access, and they were hurting for better analytics, and it just came up in a conversation.
They said, well, give it a try.
Andrew: Access meaning that database that we were taught in college is an understanding of how databases
work. Okay? In some ways it’s impressive that they could do that in other ways. It’s so frustrating to have you even had to go into that. And so you work with them. How did the first project work?
Amit: Hi. It went really well. Um, so I think when we hooked up the system for them, loaded their data and they were able to ask first question and look at the answer, the, the literally the response that came out was, our floors are on the jaw. We have never seen anything like this. Uh, so, so we knew we were onto something, but like after a.
Our team went and met with them and asked them how things are going. And they said, great, but here are a bunch of things that we would like you to do. And they came back and we used to have an entire wall that was a whiteboard. And that was filled with everything that they wanted in the products. It was, it had sort of that lightning in the bottle moment.
But it also a really immature product that had a lot of gaps that needed to be filled.
Andrew: What do you mean? What worked? And then gimme an example of what didn’t work.
Amit: Um, so it’s hard for me to believe now that we were able to deploy this product when we didn’t have the ability to, um, express, um, calculations, derived calculations. So, like, if you wanted some revenue, you could do that. If you wanted some cost, you could do that. But if you wanted to do some revenue minus some cost, you couldn’t do that.
So all of that had to be done in ETL. Um, initially, if you had multiple tables and there was different ways of joining them, we couldn’t allow that. So there had to be only one way of joining two tables. Um, so things like that.
Andrew: And then can you gimme an example of what they suggested that made you think, well, of course we need to add that we, we couldn’t have thought of it on our own, but it makes sense now that we’re working with a real.
Amit: Uh, yeah, so, um, this whole, um, multiple join paths across tables. was something that, um, we just didn’t think would come so soon. We, we thought that like, okay, some people have complex schemas, but most, most of the people will just have like, like a fact table and maybe a few dimension tables in one way to join.
And, uh, they said, nope, like we have this contacts table and this contacts table sometimes has people who refer to the customer itself. And, um, there was one more variant that I can’t remember now. Um, so there are three different ways of joining to that table. And, uh, that became fairly obvious that we’re gonna go and, uh, we’ll need to build that.
Um, calculations was another thing. We always said that, you know, if you wanted a calculation, you could always do that in ETL. But when you’re selling to business users with very little IT support, It’s, it’s a tall ask. And so having that in the product was really important.
Andrew: All right. I should say this interview is sponsored by lemon. io. If you need to hire developers and so many people now want to add artificial intelligence to their software, and if you need somebody who’s good at that, if you need someone who’s good at anything, uh, development related, talk to the people at lemon.
And if you use my URL, you’ll get an even lower price than everyone else. They got great developers, low price already, but this is even better. Go to lemon. io slash mixer gene, by the way, speaking of AI, what is it about thought spot that. Is AI driven? Give an example of how you’re using AI
Amit: Yeah. So we have always taken this approach that AI kind of is in the background, not in the foreground. Um, we are trying to essentially serve business users the best possible insights in the easiest possible manner. And we’ll use AI in every possible way. To be able to do that. So the prime example of that today is obviously the language model stuff.
So you ask a question in natural language and you may have left a lot of So if ambiguity is in the question, you may have not specified it fully and with AI, to kind of flesh it out and turn that into concrete SQL, run it, bring it back. If, if some of the things that we assumed you meant wasn’t exactly what you meant, you will be shown what was done and then you can tell the AI to change it or manipulate it.
So, so that’s kind of one aspect of it. The second aspect of it is. It’s the idea of automated insights. So if, if we have the ability to ask questions on user’s behalf, and we have a model that knows what kind of questions these users like to ask and really scalable way to get answers to those questions.
If we put all of that together, we essentially have kind of a really powerful assistant for the user who can go ask thousands of questions on their behalf and then come back with. the answers that were really insightful. And so this capability in our product is called SpotIQ. And you can figure out if some get a notification, if some metric change in an unexpected way, figure out what may have caused that change.
Like, it won’t give you causality, of course, but it will give you a lot of correlations around it so that you can figure out if one of them is causal. It can find anomalies or interesting trends in your data and things like that. Um, so that’s the second pillar of AI and the third one, which is kind of pervasive across the product, is similar to how Amazon does collaborative filtering and say people who Um, but this may also be interested in this kind of thing.
So, so we are always building a more personalized model for every user to say, like, what this user might be interested in, which data sources within this data source, which questions, which particular metrics they may be interested in, um, or if somebody else created, um, an answer, who, which other users may be interested in that.
So. Fitting that all out so that every user gets a personalized experience that’s relevant for what they are trying to do in the tool.
Andrew: I feel like a lot of that is necessary in, in AI in general. We keep thinking about how do we ask it the right questions, but some of what you said is even more interesting. How do I have an AI search or, um, ask the questions. I would wanna know, find the answers I need. And then if they’re interesting, only then serve it up.
You know what I mean? Like, I, I went to chat G p T and I and I. Uh, typed your name in and, and, uh, I came up with a bunch of stuff and I did the same thing in Notion AI’s, uh, system and I asked about ThoughtSpot, but it’d be interesting if some tool could also say here are 50 people you didn’t think to ask me about, but I think you’d be interested in and are similar to Amit and would be interesting for you.
And essentially that’s what you’re trying to do, right?
Amit: yeah, yeah.
Andrew: How far back did you add AI into
Amit: Um, so this was probably somewhere around 2016 when we started doing machine learning in our system. Um, so the very first one was we, so our search experience is very much modeled after Google. And so auto completion is a big part of it. And having the right suggestions when you’re trying to ask the question is an important part of the experience because that helps you formulate the right question.
Um, and, uh, when you, when you’re ranking those suggestions, we, we needed the AI piece to start kicking in because different users have different context and they ask different style of questions. And when someone new comes to the system, they may not know much about it. So they need to be able to leverage all the knowledge from all the other users who’ve been using the system.
So, so that was kind of the first piece. And then shortly after the automated insight piece came. Uh, so that was the second one. Uh, the conversational piece we’ve been building since 2017, but in business, it’s really important to, um, have. Never lose the trust of the user and natural language tends to be very ambiguous and the accuracy is important.
So until recently, we had a system that was like sometimes for some data sets, 80% accurate for some data sets, maybe 60% accurate. And we didn’t deem that good enough to be able to launch that in the business world. As soon as we saw the power of large language models and started integrating that into our ecosystem, we were able to push the accuracy high enough that we kind of launched that piece, um, I think last month in alpha and, uh, it’s been going great.
Andrew: What changed in that period that allows you to do it?
Amit: Um, so this, this project really started in earnest in October of last year when we first started evaluating GPT 3. 5 and, um, there was something different about, um, the level of reasoning and the level of, um, real world knowledge it had. And, um, so that’s when we kind of re architected our NLP engine to inject GPT right in the middle of it.
And, um, it took three, four months to build it.
Andrew: How do you see what you’re doing for bigger businesses applying towards smaller businesses in the future?
Amit: Um, so we, we have a range of customers starting from like 5% startups to like the. For, uh, I think three or four or fortune five companies and it’s been really useful for the entire spectrum of businesses. So anybody who is trying to be data driven, it can be, it can be really useful and powerful. And what I see on the smaller business side is.
More often, they use it for internal purposes, but they, their biggest and most powerful use case is actually embedding ThoughtSpot into their product so that they can have a differentiated data product, um, offering to their customers.
Andrew: So they could have data that’s more intuitive, easier to use than they would otherwise.
Amit: Yes. Uh, so,
Andrew: Right. Can you give me an example of a,
Amit: so. Yeah, so, um, I’ll give you, uh, there’s a series B company called clear now. And, um, they help. Importers and exporters, um, get custom clearance. So, so they have digitized that whole experience, what used to be paper and pen kind of experience. And, uh, initially it was all about just being able to fill the form online and, uh, do the transaction.
But pretty soon their customers started demanding. Um, data from their system, like how many shipments have you cleared, what percentage on an average, on the value, am I paying customs, what are the broad categories of products and things like that, and. They, they were at this fork where either they needed like a five person team constantly responding to new kinds of questions and building new dashboards and embedding that into product or enable something that’s very intuitive and self service that they could expose to their own customers.
And that’s when they decided to choose ThoughtSpot and embed it in their product. So that’s kind of one example, but there’s several similar examples where. Um, product companies are trying to figure out either how do they have a differentiation, uh, with, um, and their product versus other products and how friendly it is to get data from them, um, or trying to monetize their data in a way that’s conducive to consumerization.
Um, or, or maybe it’s just like, um, a necessity and they want. to spend as little engineering and data teams bandwidth as possible and still be able to serve their customers.
Andrew: Um, when I looked you up on LinkedIn, I saw that you were very proud that a small team, five to eight engineers, was able to add hundreds of millions of dollars of incremental ad revenue. You now have a team that’s huge, hundreds of people. How’s the transition been? Going from a small team to something that’s this big.
Amit: Yeah. So it’s a very interesting journey every step of the way and a lot of learnings. Um, when I left Google and we started ThoughtSpot, pretty soon we had a team of Um, four or five people, um, so similar size team, but in the beginning, all this wonderful infrastructure that I used to live inside and Google was gone and we had to create it from scratch.
And, uh, that was almost like. Similar experience to when I moved from India to US and you don’t even know how to cross the street anymore and you have to learn that it’s kind of similar thing that like, how do you build your code? How do you, um, where do you set up your get repository? Is it going to be online or on a desktop under your desk?
And all these micro decisions that were made for me in a large organization, we had to figure them out and. Knowing that it will have large implications, uh, down the years. Um, and yet, sort of, you didn’t have much data and you had to make quick decisions. So that was very interesting. Um, being in a big company, I never had to worry about selling the job to candidates for the most part.
My role, at least where I was at Google, was mostly around evaluating people, like giving them Challenging problems and puzzles in the interview and saying whether they are good enough to join the company or not. But when you’re a startup, the much bigger job is to convince somebody talented that this is a bet worth making.
And, uh, that, that’s, um, that taught me so much about people, about business, about just having, learning and upping my game on communication. Um, there’s a lot more of dealing with ambiguities and you just kind of, one of my mentors once said that you never make the right decisions, you make the decision right.
And that was kind of, um, aha moment for me that like, okay, that’s how you’re supposed to run in this ambiguous environment. Um,
Andrew: Yeah, I did see on one of the earlier versions of your website, there was a tab on the right side of every single page that said, Earn 20K. I said, what are they doing? Is this a marketing thing? Nope. It was help us make the right hire. I think you even said you’ll sponsor visas for people who are currently on an H 1B visa.
Here it is. It was such a critical part of your business in the beginning. So, how did you convince people to, to bet on you?
Amit: Um, so I think, um, a few things that helped was, um, we, we had a reasonable Reputation in terms of, um, Ajit was coming out of co founding Nutanix, which was already at the time by 400 million in last valuation. Um, I was coming out of a very successful run at Google. We were backed by Lightspeed. So, so those things definitely helped, but ultimately we were trying to hire.
You know, who are best geeks and what really convinced them was the, the grandness of the vision and, um, the grit with which we were attacking these problems and they’re like, this is an exciting journey. I want to be part of it and I like these people and if I had a bet on anyone to do these kinds of ambitious things, these are the people that I would bet on.
Um, just getting that across was, uh, probably the most critical part of it.
Andrew: If you weren’t doing this, what are some businesses that you would start today?
Amit: Um, so, as I mentioned, um, early on, my childhood dream was to go be a professor somewhere and do fundamental research. And so, I kind of keep dreaming about… Um, how, how do I do work that contributes to expanding the envelope of human knowledge? And, um, you know, if I wasn’t doing this, I would probably want to jump into, um, building a research lab like DeepMind or OpenAI or something like that.
And that’s where. I think I would have a lot of fun if I was, I wasn’t doing Taskbar.
Andrew: That would do what? Like DeepMind, basically having software take on the big challenges, the big problems of the world and see how it could solve it.
Amit: Yeah, so I, I really find the vision inspiring, which is that, um, build AGI so that you can solve all the other problems in the world in a much more intelligent way.
Andrew: Interesting. What’s your schedule like? I’m trying to get a sense of how you operate. How do you work? Do you have like a set time of day that you work? I was just talking with the founder of clear a bit. He says that one of the things that he does is he journals all the time as a way of organizing his thoughts and becoming more productive.
What is it that makes you who you are and makes you as productive as you are?
Amit: So I don’t think. I am probably the best example of organizing time. Um, I, um, at this point I’ve managed to, um, create a job for myself that I absolutely love. And, uh, it involves very little of what I don’t want to do and everything of what I want to do because we have amazing leaders running engineering and operations and business and everything.
at the company. Um, so my days are mostly focused on one of the two things, either, um, helping my team make progress in directions that could be potentially big arcs for the company. And these are sort of small experiments, large experiments, or just kind of breaking our heads together in solving a specific Many problem that is on the path to solving a much larger problem or, um, just, um, helping communicate the vision and the technology and the product that we’ve built to the outside world, whether it’s customers, prospective customers, public at large and, um, and the third piece of it is just, I feel like, Being a constant learner is critical to what I try to do for the company, so I try to reserve some time every day, maybe an hour, maybe a little more, just reading and learning and talking to people who know more than me about a particular topic.
It’s, uh,
Andrew: And so you’ll just sit at the office with a book or with, or go online and do your research, understand this one topic and that’s part of your day. Mm
Amit: yeah, I mean, it usually happens at home, but, uh, most of the day I would be In meetings and these meetings could be like a whiteboard with two engineers or it could be like a broader group where we’re trying to figure out how do we launch this product and what steps are sort of along the way or with a customer just sitting on my desk reading and writing or just calling up an expert in a particular area and trying to understand the lay of the land for a particular thing.
Andrew: What’s a topic that you’ve tried to understand that doesn’t directly have impact on your work, but you just needed to know it
Amit: there’s a large range of things that I’ve been reading off late that’s not related to work. Um, some of the literature coming out on longevity and good health has been really interesting for me. Um, like David Sinclair’s book, um, there was a, it’s, it’s not exactly related to my work, but it fascinates me.
Um, There’s a book by Jeff Hawkins called Thousand Brains. He’s dedicated a big part of his life in understanding how brain works. And I always find his writing fascinating. So that was another book that was not related to work that I read.
Andrew: and David Sinclair’s book is lifespan why we age and why we don’t have
Amit: Yeah.
Andrew: and and so do you think that this? How are you applying this yourself because you just don’t want to grow old the way that people did maybe in the 50s Or is this your belief that we are on the cusp of somehow expanding life?
dramatically
Amit: So part of it was intellectual curiosity, because I’d heard enough about it from different people. But the book definitely inspired me to live a lot healthier than I was living. And so, this was kind of a turning point for me in terms of doing intermittent fasting, increasing the amount of exercise I’ve been doing.
And I, I’m not particularly interested in living to be 150 or something like that, but I’m definitely interested in increasing the health span and being fit independent person till, like, maybe late 80s or something like that.
Andrew: Okay, all right right on the website is thought spot. com. You still write your own blog posts on there
Amit: Ah, yeah, I do.
Andrew: They’re
Amit: I do. So, um, sometimes I would write on Substack. Sometimes I would write on ThoughtSpot. com.
Andrew: Yeah. You’re pretty prolific. Sometimes on LinkedIn. I think I’ve seen you. Yeah. All right. Well, thank you. Thanks for doing this. And as I said earlier, if, uh, you’re looking for the next podcast, I’m doing a podcast on Dow’s decentralized autonomous organizations, and I’ll tell everyone in the audience and you too, Amit, um, and your team, you’ve got the team here with you, uh, watching and making sure it all goes well.
I’ll tell you that my new podcast on how these organizations function is available at. Join origami. com slash podcast. Thanks. Bye everyone.