Transcript: Satya Singh - Co-Founder of Scispot on redefining modern biotech's data infrastructure

Host: Hey Satya, welcome to startup project.

Guest: Yeah, thank you. Rathraj excited to do the podcast.

Host: Uh, so, you know, we've been, we've known each other for a while. Uh, we met at this dinner that Martin hosted, which was quite an interesting dinner, I have to say.

We had some very interesting people in that dinner and uh, I think we are, I think only we two were Indians, if I remember and there was one other guy or two other people who were probably like CEO's of um ex CEOs.

Um, so yeah, it was a great dinner and you know, ever since that, uh, I've been following Saispart and the interesting work you're doing here.

Um, so I thought it'd be a good uh conversation to have on the podcast to have you on and talk about your experience. Uh, just to set the stage, uh, you know, can you start with um what's your background?

Where did you grow up and how did you sort of uh get into tech?

Guest: Yeah, um for sure. I think the dinner that you mentioned uh was definitely uh really insightful.

I met some really interesting people uh who have kind of built businesses over the last few decades and shared some of the lessons learned as well.

In terms of like my background, um I grew up in Agra in India, which is also known as the city of Taj Mahal.

Um I always had an option to uh come to the UK because my uh grandfather was like part of the British government services when uh British uh were in India.

So uh moved to UK really young and most of my career journey has been in the UK, but obviously currently I'm based out of Seattle.

Um my passion has always been like data uh right from the beginning and I guess it's stemmed from like um it's a funny story, but I'll share it.

It's stemmed from like learning numerology when I was 12 years old and I learned a sign language where I converted every alphabet into a number. Even now like when I see alphabets, I'm just like converting that into numbers.

So that kind of drove my passion for like numbers if you will and that's how I got into data. And um yeah, started my journey with Virgin Atlantic, uh worked for Lotus Formula One, worked for consulting, oil and gas.

Um also did my shoe business in London, uh so did a fashion startup. Um and then worked for Discovery Channel, Eurosport Player uh doing data engineering, data science work, product.

And then uh landed in Expedia uh in Seattle and um yeah, so that's been kind of my journey before Saispart if you were.

Host: Tell me a little bit more about um Jackson, right? This was your uh startup. Uh what were you doing with that?

Guest: Yeah, so I love boots, um you know, men's boots and um I used to design boots in my spare time.

And at the time I was working for hotels.com leading their growth marketing team, so I had a really good idea on like how search engine marketing works, how search engine optimization worked.

So I uh I got some connections in India uh with a shoe factory um because I mean whole of my family is in business. My dad, we three brothers, all of us are in business. So I had that edge to like start something in shoes.

So I started Jackson.co.uk and uh kept it simple to begin with, like B to C uh with a website, uh running a lot of campaigns and I was selling like 20 pairs shoe shoes a day, which is not a lot, but for me like starting from scratch, that was quite a lot.

And then I kind of reached out to mom and pop shops up and down the country uh in England and Scotland and started selling B to B shoes as well.

And I pretty much designed all of the shoes and uh the audience that I was targeting was like 40 to 60 year olds who look for comfortable shoes, but still want to like have stylish boots uh in the UK.

So it was quite successful uh but I ran into like supply chain issues. I didn't realize that certain sizes would sell out sooner than later. Should have run a data science model based on the survey uh but lesson learned.

And uh basically sold that business. Um, but it was, it was amazing. Like I had it for three and a half years and I I loved every part of it.

Host: Was it a a good exit for you or what was that exit like?

Guest: It was a good exit, so I sold it to like an existing company in the UK, which also bought my uh stock that I had. uh So it was a good exit. It wasn't like anything like what you have would have in a Silicon Valley exit, but for like a mom and pop business, uh kind of it was it was a really good exit. So I got a lot of savings out of uh selling that business.

Host: Nice. Um, so you moved to the US uh because of Expedia?

Guest: Yes, so Expedia kind of promoted me to a director position. So that position was in Seattle. So they moved me from London to Seattle to lead uh their platform team, uh so security and data platform team.

Um, so that experience was like really good because obviously I was working with uh the US folks, but still there was a slight change in culture compared to like what the London working environment is.

But I thoroughly enjoyed uh leading a team of like security engineers and data product folks.

Host: Got it. So at what point uh you made this transition, I mean you have a nice cushy job in Expedia. I'm assuming it's cushy. Um, but at some point you said, you know, uh this is enough and you started uh Saispart. Uh, why the transition and how did that happen?

Guest: Yeah, great question. So we're three brothers.

I'm the oldest, although I would, I would say I'm the youngest, but I'm the oldest of all three and the middle one's already doing his business and the youngest one at the time was working for an AI lab based AI company in Germany in Berlin.

And um he spoke with like 300 life science companies as he was commercializing uh their efforts and one thing he realized was um there's a huge gap in the market when it comes to data products for life sciences.

So he moved to Canada, started Saispart. At the time I was working for Expedia. Um, so he pitched the idea to me, my brother, um and the reason I like jumped, jumped uh with the idea was A like I've done business with this brother.

Growing up, we used to like handle our dad's hotel business in Agra, so we kind of did business together, so I knew his skill set and had, you know, like working with brothers can be hard, but for us it was just like we've done this all our lives, it's kind of natural to us.

Um, so I loved the idea and um it was hard to explain to my family uh my wife like going from a really high paying job to like starting a business. Um, but after explaining the reason, uh she was also bought into it.

So I kind of left my job, started Saispart and then luckily got into Y Combinator the next month. Um, and yeah, and then we never look backwards, it's been a great story so far. We grew right from the first month even without product.

We had a customer, you know, who helped us develop the product. Um, and since then we've kind of grown quite a lot.

Host: So uh what was the pitch to YC when you know, I think it all happened pretty fast, right? Your brother pitching to you and you convincing your family and then joining YC. So what was the pitch to YC when you were first trying?

Guest: It's funny like at that point in time our pitch was slightly different to what we ended up building.

At that point in time we wanted to build a conversation platform because I was actually like uh working with an Expedia uh with the conversation platform and building security protocols around it and um and LLM's weren't like uh as famous as they are right now, but our initial idea was like Cybot, uh Cybot AI, which basically like helps you uh find samples using voice, you know, using AI uh within the lab.

That was our like initial pitch. But then um that evolved, we pivoted and we got feedback from YC partners. Um I think we were ahead of our time essentially.

I think we now we kind of get into the conversation part of it really easily, but um what kind of resonated with our customers were like data orchestration, the data platform connecting the data from instruments.

Um thinking of data as a product because a lot of experiment data is unstructured. Uh so building like an unstructured data lake that connects the experiment metadata with instruments.

So we got clear signals um from the market and that's how kind of we uh developed Saispart.

Host: So what was exact problem right now uh Saispart solves for biotech companies or life science companies?

Guest: Yeah, so the main problem is that 80% of biotech data never gets analyzed and there are various reasons for that. One is like um you would be surprised, a lot of labs are still paper based. Like in 2024, uh there are paper based labs.

Doesn't make sense, but there is. Even like some of the enterprise companies have some of their processes in paper.

And the second reason why 80% of the data never gets analyzed is um the file format and the sources of data is like quite unique in biotech.

For example, you might be working with an instrument that has binary files and no kind of horizontal product would be able to like interpret those files. You know, even with foundation models, the accuracy might not be great.

And then the third reason we've seen is within the lab there are a lot of handshakes that happen, you know, so there are like very like technical wet lab teams, the very technical dry lab teams who uses orthogonal tool sets and they don't like do like handshakes within one platform.

Um, so because of those three reasons, uh as we did the market research three years ago, uh we found that 80% of the data is like not even like utilized at all.

So that was like the impetus to start Saispart uh that how can we help these labs not just like connect the dots, but reduce the noise to signal ratio as we connect uh different data points from instrument to experiment, connecting the wet lab to dry lab, for example.

Host: So you're are you a vertical sort of SAS solution uh to labs?

Guest: That's correct. So um we are a vertical SAS solution for biotech R&D, manufacturing, quality uh different types of lab based businesses.

Um, and what we're seeing is like biotech will be the largest producer of data uh even surpassing astronomical data by 2025.

So that's the trend we're betting on because we want to make sure we help these companies, especially modern biotech companies to be able to utilize uh the power of LLM's and I think often there's a lot of glamor around foundation models and LLM's but people don't really talk about the grunt work behind the scenes.

That can help you leverage those models.

Host: I mean, I just want to go a little bit back to YC. Um, how was the YC experience? And sort of what were the big takeaways of that experience for you guys?

Guest: Yeah, so um it was 20 uh YC summer 21 and uh it was all like COVID, everything was remote. Uh, so it was like a very interesting and unique experience, I would say because we didn't get to like uh go to SF, meet uh our partners in person.

Um so it was pros and cons.

I think the the cons were like we didn't have that collaborative environment where we can meet in person and learn from the folks uh there itself, but the pros were we we learned how to operate in a remote environment completely as as building a remote company, we've been successful.

So what we learned was how to build trust virtually, you know, how to like uh run uh zoom meetings properly.

I know these sounds very trivial, but we learned a lot of like uh soft skills uh during YC, but But yeah, I would still prefer going back if we could have more in person meetings. I think that would have been helpful.

Host: Other things that I mean, I know I mean YC is such a big brand name and uh with you know Sam Altman who used to head YC before Gary Tan, you know, it has all its positives. Um, what are there any things that people don't understand that are really valuable about YC or like the things that you get out of YC that people don't, I mean what makes really YC YC?

Guest: I would say like the biggest lesson that I learned from YC is like being very humble and open to market signals. Like a lot of founders have this kind of passion that their idea is the best idea and it will work no matter what.

Which I think is a good place to start with, but what YC taught us was, you know, ultimately market dictates, you know, what the output or the outcome would be and really listen to those market signals and not get too attached to your idea, you know?

And I don't think like people understand that from the outside who've not been part of YC.

Um because our partners were like they asked such good questions that they helped us not get too attached with the initial idea, but get attached to like the market signals, the the data that we're getting, the anecdotal feedback that we're receiving.

And I feel like Paul Graham has says like if you've seen some of the essays, he's written are so good. He talks about um doing things that are not scalable in the beginning, you know?

And a lot of people and they like who've not been part of YC always want to like build this growth model that you know, grows exponentially, but you got to start things that are not scalable and really understand the market signals.

So I get I guess getting trained on that was was really good during during YC.

Host: Yeah, I completely agree with that. I often like see pitch decks and whenever someone asked me, hey, is this pitch deck good or what we should do about, I always tell them, you know, your pitch deck should be your truth seeking exercise and not like what your assumptions are there and you put that in the pitch deck.

Guest: 100%. People always try to put their own ideas into the pitch deck, but not actually you know, look at it as a hey, this is a truth seeking exercise and market might have a different opinion about what you are thinking.

Host: Do you think sometimes like going after scale kind of disrupts that truth seeking exercise?

Guest: Going after what? Scale? Some sort of scale or exponential growth?

Host: I think especially because I see a lot of preceed pitch decks, um I think sort of people want to fit in that venture mold because they're making that pitch, so make they'll make these lofty assumptions and create this whole story that might not really exist out there or that you'll only know when you actually go through that path, right?

You can only make some assumptions, only events you go into the market, you'll really know what the truth is. So you have to be a little bit flexible in terms of uh I mean in preceed, you know, some founders come with financials.

Although like yeah, it's a good exercise, but we both know that it's not going to work out the way we think it is going to work out, right? Um, yeah. I think the loftiness for sure adds uh that distortion.

But I think the loftiness is a good excuse to be more truthful and find the really big idea than sort of going towards a small idea that you have and expanding and force fitting it into something it isn't, it is not, right?

Guest: I agree with the big idea.

I think the part that I have problem uh understanding is just just highlighting the tam because best founders have created their own tam and we've got plenty of examples of that as well and I feel like sometimes founders and I've been I've done probably the same in the beginning stages is like we get over indexed on like how big is the market, you know?

And I feel like as you get market signals, you can define your market that is uh uh eligible for venture funding, you know?

Host: Yeah.

Guest: And it's also like tam is just I, I always tell people that it's just a Boolean signal. Is it a yes or no? Is it big enough or not? It doesn't matter really exactly what number it is. Like is this big enough opportunity to pursue or not?

That's the only question you that tam slide should really address and nothing more than that. Um, but coming to uh Saispart, uh, you said you already had a customer going, um, you know, within a month or uh or before you had a product.

So how did that happen and how did like those first five to 10 customers uh uh came to Saispart?

Guest: Yeah, so my brother has always been in the life science industry, so he already had some social capital built in, which helped.

Host: What was he doing there in I mean life sciences?

Guest: So he commercialized uh life science uh software companies and was head of growth and head of marketing. Uh so he worked for scientist um in San Diego or science exchange in the Bay area, Lab Twin in Berlin in Europe.

So um all of his career has been working for uh life science software companies and he understood the market uh really well. Uh plus like um also really figuring out who are ICP's uh in the beginning based on the market signals we got.

So we went after them, understood what they're looking for. Uh we did some podcasts as well in the beginning uh with like biotech CEO's. And some of our like initial angel investors were like from biotech.

So um all of that kind of helped us hone down into um finding a few collaborative partners uh even before we had the product. Um, and that helped us kind of build Saispart and you know, where we are right now.

Host: So right now what does the traction look like at Saispart?

Guest: Yeah, so we're growing uh quite fast uh over the last three years. Uh we have like startup, scale up, enterprise customers, some public biotech companies as well. Uh got customers in North America, Europe, Africa, Southeast Asia.

Um, so yeah, grown up quite a lot since then. Uh initially we were like focusing on R&D uh type companies, but now we support uh different uh companies within the lab based businesses from manufacturing to quality, to R&D, uh production.

Um, and all of them kind of leverage Saispart to connect their data uh whether that's instrument data to experiment metadata to results data, all in one place. It's almost like coming back to like science in one spot, uh size.

Um so yeah, grown quite a lot since then and our focus now is uh leveraging like um AI not as like a wrapper that you know, uh I hate building chat GPT wrappers, but uh my focus is like how we can like um make it more like agent based.

So essentially science is hard. Science has more variables than traditional tech, for example. So one of the things I'm really excited about is how can we build um different agents that can delegate tasks to each other.

So for example, one agent that brings data from instrument and really understand the context of lab instrument data, like let's say QPCR data.

Uh then one instrument that understands experiment metadata and these two can talk to each other uh autonomously with some some human feedback.

It's almost like building a future smart app uh which I think the industry will head where you have bunch of agents talking to each other. Obviously there's human in the loop.

But because science has so many variables, I think uh this could be hugely impactful.

Host: Yeah, I keep seeing this line now that, you know, we don't have to talk to each other. My AI will chat with your AI and we all can go and sit in a beach somewhere. Uh, so I guess we'll reach that point.

But talking about AI, I mean before even that, I think if you're dealing with a lot of unstructured data, I think that's sort of like a very, very sweet spot to be in any sector that you are right now.

Uh, and because and to just to clarify like the difference between unstructured and structured is, structured is all the data that you put in databases and unstructured is pretty much all the data that uh, you know, that is in the form of files essentially.

And I deal with unstructured data on a data basis because I run a files is a unstructured storage platform.

Um, but I feel like you are really at the cutting edge if you if you store all the unstructured data in life sciences at one place then you become really valuable as a company or a product. Uh, right?

Guest: Yeah, our goal is to like um help biotech companies build their own IP. Like that's that's our goal. So it's almost like us not capturing the data, but be the staging lakehouse that our customer used to build their own IP.

Uh that's more of a goal, right? So unlike competitors in this field who um wants to build like a single source of truth, that's not what we aspire to be.

What we aspire to be and what the reason we've been successful with modern biotech companies is we provide the tooling for these data science focused biotech labs to be able to build their own IP.

So for example, you talked about unstructured data and with foundation model, I think it's it has become really easy to build an unstructured data lake uh with embeddings and vectorizing uh and entity recognition of some of that uh data that you uh get from like different sources whether that's experiment protocols or instruments.

So our goal is to like um make these labs successful by bridging the wet lab and dry lab gap, so they can also build their own IP as they gear towards like filing a patent or they gear towards um building a platform uh for uh drug discovery or whatever workflow they're working on.

Host: Got it. Yeah, I think you're almost describing data bricks like a platform but for modern biotech.

Guest: That's right.

Host: Right. Um, so I mean you've raised, I think couple of funding grounds. Uh how how has that journey been? Like I know like YC round might be a pretty straightforward like once you know, YC accepts, um you're obviously guaranteed one round and post YC, how has it been? Uh talk to, talk to me a little bit about how the funding scene has been in you know, biotech software.

Guest: Yeah, um it's been it's been really good.

I think one of the things that has helped us is we've we were profitable after year first uh and has been like break even/ profitable depending on how quickly we keep adding uh team members to our team.

So uh being profitable has helped us uh secure funding even in the last year climate when the climate wasn't good. Um and our goal has been to be default alive as YC says.

So always been like focused on irrespective of the funding environment have uh runway which is infinite uh if you will and um really making sure our fundamentals are correct because one thing I learned during YC days uh is there's no correlation between funding and uh company being successful, you know?

So uh YC did this analysis and it was like quite uh mind blowing to see like there's no correlation. You could be the you could be raising lots of money.

Guest: Exactly.

So our goal was to like right from the beginning, how can we build a really fundamental business that understands like build something that customer needs rather than uh what I would call like a sysy business like solution in search of the problem.

Um, so I think that has helped a lot. And funding hasn't been our metric that we ever optimized for. Uh, what we've optimized right from the beginning is like happy customers, you know? Can we create a success story with this customer?

Um and if that's a fit, then we work with that customer. So we've been very picky as well in terms of onboarding customers and you know, creating case studies with them.

Host: All right. And how how long does it take? I mean talk to me a little bit about like a your approach on getting a new customer and how long does it take from you know, reaching out to a customer to like you know, onboarding them to as a customer officially?

Guest: Yeah, so I lead product in tech, my co-founder and CEO uh and his team handles uh sales and marketing, but at a high level, um we have a lot of inbound, so that drives a lot of growth for us.

Referrals and inbound from our existing customers who kind of um get us more customers through word of mouth. Plus marketing, uh most of our channels has been organic so far. So we've had a lot of organic growth.

Um, we're also rated as like um the the easiest way uh in terms of like UI UX of using a platform within this industry from G2.

So uh we were ranked number one from our customers in the last quarter, so that has also kind of helped in terms of the inbound.

And um in terms of like onboarding the customer, um depending on the stage of the company, it can take anywhere from like two to six weeks because a lot of companies are like work in regulated environment. Compliance is really important.

What we call like GXP control environment. So um two to six weeks is I would say like from the day one to getting the customer onboarded, which is the lead cycle time is very less compared to like other competitors in the industry.

Um, and the reason we've been successful at reducing that uh cycle times is how we build a platform because most of the biotech software companies started as like data management platforms, right?

We started as like um data orchestration and how can we orchestrate different workflows without introducing tech debt uh within the platform, right? Because we we are lucky that when we started, we had that technology, right?

If he started a similar company solving a similar problem, let's say 10 12 years ago, um when the volume of data wasn't very high, it was all about digitization, right?

I feel like digitization problem is slowly going away, although it hasn't completely gone away and the focus now is like how can I like orchestrate the platform, uh build workflows on top of it based on uh my specific needs, based on like drug discovery or whatever research that I'm doing or manufacturing I'm doing within the lab based businesses.

Host: Uh, you were you know, doing product at Expedia, right? So how did you know, that help in what you're doing, you know, in product development at Saispart?

Guest: Yeah, so I think in travel tech margins are really low compared to the life sciences or bio industry. And as a direct uh result of that, even competitors kind of share API details and collaborate.

Um so one thing I learned in travel tech being product was there's a lot of like openness in the ecosystem.

Uh for example, there's hotels.com API, there's Trivago, there's Trip advisor and everybody kind of bids on their specific uh cost per click model to understand where they are in the business and then share profits depending on if I'm bidding on business travelers versus family travelers.

And I feel like bio uh or life sciences is still far away from that model. Um One thing that I would love to see is my experience in like data products and API's uh helping the ecosystem in bio and life sciences to be a little bit more open.

I know in the regulated industry, compliance is the biggest reason, which makes sense and um but having said that, you can also securely collaborate with indirect competitors or do more partnerships that can help um not just from a profitability standpoint, but also in terms of the long-term goals, whether that's bringing drug to market or whether that's launching some product in the market, uh you know, uh if you're manufacturing proteins, you know, for instance.

So I would love to kind of have more API uh secure API's to build that partner ecosystem in bio.

But how that helped directly is uh from a data perspective, I did feel like I've gone back 20 years though uh in terms of the data products I've seen, but that makes me more excited because I feel like Saispart can have a huge impact uh in terms of building data products in the industry.

Host: One of the challenges I hear when you're doing early stage is, especially when you have continuous stream of new customers is how to prioritize what you build next? Because you know, customers have one off asks.

You know, that sort of puts a challenge in terms of what you're going to build next.

So how, how do you like do you have a framework that you sort of uh think about on what you're prioritizing and because you have to both balance new customers and also balance growth and then balance business priorities like a give me a, you know, your thought process around that.

Guest: Yeah, um so that's that's a harder problem to solve. I wouldn't say it's an easy problem. We battle that on a daily basis.

Um, but one thing that has helped me is I have to unlearn some, I I had to unlearn some of the things I learned in a bigger company. So I worked in both startups and bigger companies before Saispart.

Um, and the bigger companies taught me like a framework of effort versus value having a business outcome. And I feel like some of that doesn't apply in startup. Um, meaning it's not always effort versus value.

It it has to tie with um your ICP, what kind of persona you're targeting. Um, so we we have a repeatable sales cycle. So we understand at least for the next two years what kind of persona we're targeting.

So we can set up the company, the customers for success and have a really good case study with them.

And based off of that, uh the top down approaches to uh prioritize the features because it aligns with our roadmap and plus it would help us, you know, um create a good success story with our customer.

So that's like one kind of signal that goes into the road map. But it's still very challenging because tactically that could uh still not fit into your effort versus value framework, right?

Because some of the value might be higher, but effort might be higher too. So the other signal that we put in is uh making sure there's no technical debt as we are building our platform.

There's always some technical debt, but it has to be very minimal as you uh build your platform because if you do things right, your growth might not be exponential, might be linear, but what that would help you do is when suddenly you explode, your platform would be ready for it, right?

So we've always had that approach of like the orchestration and um having technical debt almost zero from the beginning. So that also drives um the the road map. And the third signal that I have is the security and the infrastructure.

So I led the security team and um you know, understanding like sock to hippa and all of those controls was fairly easy for me coming from uh security background.

So making sure that whatever infrastructure we build so far, it's easy to like satisfy all the controls in real time. Does that make sense? So like it's not purely versus value, but it's more complicated in that.

Host: Yeah, that makes sense. Um, I want to ask uh how do you see Saispart in the next couple of years, you know, in the short term one or two years versus you know, maybe in four five years.

Guest: Yeah, in the next couple of years, like our focus is to be the staging lakehouse. So how we're currently helping customers is irrespective of what tooling they use, uh we fit into their stack in an interoperable way, just like a middleware.

Uh, so they can connect with their existing tooling stack because there's a lot of tool fatigue and we don't push our customers to have hey, Saispart is like your one source of truth or Microsoft for bio. That's not our approach.

Our approach is to be that middleware. Um for two reasons. One is so they can build their own IP and leverage uh some of the foundation models.

And second reason is so they can be more compliant and have chain of custody for all of their data, meaning um the handshakes they do from like instruments to wet lab to dry lab to results to reporting uh it to FDA or um preparing like compliance reports.

So those are the main outcomes that we're targeting in the next couple of years. Uh long term, our goal is to like um let's say if any of these companies want to um build IP and uh file a patent or build their own graph databases or ontologies.

So in biotech there's this term that is quite famous called ontologies because everything is so like many to many connections. Uh think of a sample, you might derive like aquits or derivative from that sample. Then you might connect results to it.

Uh then you might have some deviations. Uh so you're doing same thing over and over again with some protocol deviations.

So um building ontologies or building a knowledge graph, um every company can kind of build their own AI knowledge graph using Saispart API or uh Saispart graph databases that can then be part of their IP uh is what our long-term goal is essentially.

Host: Okay. So we are almost at the end of the conversation and pretty much all the guests I ask some common questions, so before getting into the audience questions, I want to ask those with you. Um, what are some of the favorite books that you've read or are reading right now?

Guest: Yeah, so like I read zero to one, which kind of taught me, you know, um how to think of a startup uh going from zero to one and also like I've worked in startups and in bigger companies.

So I've gone from one to end, but there's like a really different uh skill set that's required from uh going from zero to one. Then I've also read a lot of things from Paul Graham uh online that also helped me think more strategically.

Uh Peter The has also been another kind of you know, thinker that has influenced me over the last few years. And then recently like um read a lot about uh network effects.

Um this book from Andrew um I forgot the name, but it talks about like how a startup builds network effects without having a network, you know?

Host: Is it Andrew Chen from Andrew? Yeah. Do you remember the name of the book? Well, I've read that book like six months ago.

Host: I I started reading it. I think I was in midway and I stopped. I mean it sort of went away, but yeah. Yeah, I remember that book.

Guest: Awesome. So I I love that book as well. And then I'm also reading a lot of books in the the bio domain. So I come from the data background. My brother is a molecular biologist.

So I'm learning a lot like one of the studies from McKinsey states that a lot of businesses would become lab based businesses in 10 years. Um, like we will be manufacturing a lot of different things in labs, right?

And what is that trend and how does how will that change humanity?

So I'm reading a lot about that and learning as well because um there's a lot to learn in the in the bio domain and what kind of keeps me going there is we've made a lot of pro progress when it comes to like um digitization or um AI for instance, but we've not made much progress when it comes to understanding nature and the variables.

Like you know, there's so many variables in nature. It's like thousands of years of history or models that have evolved over thousands of years, which So that part really kind of fascinates me.

Host: Yeah, I mean to actually um Kevin Scott, who is the city of Microsoft in his strategy or sort of AI prediction for next couple of years, what he stated was AI being used in biotech to invent new things is going to be a big trend uh and how it would also increase the need for computation and at the same time the time required to do this innovations will be reduced.

So I think it's overall an exciting trend to watch.

---