Sumon Sadhu: “The Portuguese are really good problem solvers”

May 21, 2016

The Lisbon Investment Summit 2016 is getting closer and to tell you the truth, with so much that’s going on, so many awesome speakers who are coming to Lisbon, I can’t really tell who from our team invited Sumon Sadhu, top investor and founder of Muse.ai, to speak at #LIS16. I just remember sitting in the office the other day while Isabel Salgueiro and Inès Dartiguenave from Beta-i were telling me that I had to interview him because he had a very interesting investment philosophy as he only believed in working on things that will make a deep and meaningful impact on the world – often going against permission.

That short description was more than enough to get me excited about this interview. So, I took this suggestion and had a 40 min chat with Sumon where I pretty much just listened and felt like Jon Snow (I really knew nothing about artificial intelligence, big data or how to extend humanity’s lifespan). Sumon really has an exciting story full of minor funny details that make you wonder if anything in life is often due to chance.

The man who moved from a promising career in science in the UK to be an entrepreneur in Silicon Valley, because Paul Graham from Y Combinator misunderstood his business idea, has really a lot to tell. But, I’ll let him explain it through his own words in the interview below.

Why did you become an entrepreneur?

Because it was the only thing that allowed me to express my ambition. So, I was originally a scientist, and one of the things when you have an ambition and you’re a scientist trained in biochemistry, you normally want to publish a paper or cure a disease. And that’s a very noble goal, but the reality of being a scientist is that even though you have big ideas and big aspirations, which keeps attracting people to science, is that the end result doesn’t come with the same freedom as if you’re an entrepreneur. If you’re an entrepreneur you can imagine something that you want to do or something that you want to change, and there is no one but yourself to stop you from really doing it. So, being an entrepreneur comes down to taking things that you imagine and bring them to the world, without asking for permission.

How was it like having your first startup, Snaptalent, and going through Y Combinator?

At Snaptalent we were essentially students at Imperial College London and we observed the way that the world worked had sort of changed in terms of behaviour of young people online. And when Snaptalent came about, Paul Graham from Y Combinator reached out over the summer and the email he sent me essentially changed my life because it created a new opportunity. The idea we applied to Y Combinator with was very different from the idea that we did Y Combinator with. So Snaptalent was prompted by a conversation we had in East London. Paul Graham likes to reason about ideas and he misinterpreted something that we were saying. He ended up saying: “What if you guys built the adsense for jobs?”. So, we then explored that and built a prototype and brought that idea to market. Moving from the UK to having a fully funded company in Silicon Valley gave us the permission to do something – Paul Graham kind of gave us the permission and legitimacy to believe in yourselves and go after this. I dropped out of my Phd and we built Snaptalent through the duration of Y Combinator and started advertising. We made 30k during this period and we ended up raising 1.5 million dollars after YC. But it was a remarkable transition of us not believing in our ability in Europe and then being told like ‘you guys should do this and I give you permission’. For me this is the interesting relationship of permission with being able do something.  

You ended up shutting down Snaptalent. What were the main lessons learned?

Well, first, it’s really important for you to understand your co-founders in crisis, as well as in success, because people’s response in crisis is often different from their response in success. So, that’s the first thing. For us as a team when the chips were really hard we wanted to go in different directions. Another lesson is around market timing. Some things cannot be changed and market timing is one of those things. You can’t build a product that it’s too ahead of its market. There’s just too much friction to bringing something like that to existence. For example, building Snaptalent in 2016 would be very different to building it in 2008. In 2016 most of the impressions on the web can be purchased on an auction basis, there’s also increasing amounts of data about users. And so the funny story with Snaptalent is that in 2013 I became an advisor to a company called Worked For, which is now the top recruitment solution based out of France, and they were literally saying ‘we learned about what you did with Snaptalent and we want to bring this back into the market’. So, market timing is absolutely critical and it’s just one of those things you cannot change as an entrepreneur.

What about Quid? How did you end up joining them?

It was right after Snaptalent. I had actually been advising the guys from Quid because we were friends from Oxford University. That’s the big thing about my career, it has been very held to relationships. Bob Gudson, the founder of Quid, was actually the first person from the student entrepreneurship societies we created in the UK, to move to the Valley. He was head hunted by Max Levchin, the founder of Paypal, to join his incubator, which had companies like Yelp or Slide. I had known Bob for while, we were friends and we would help each other out with our ventures, so when we had to shut down Snaptalent I asked Bob if I could help out with Quid. I also needed to get a new Visa because in the US your Visa is tied to your company, so if your company runs out of business you have to go back to your country. So, at the time, the important thing for me was to stick around in the US and for that I needed to join another company. I told this to Bob, ‘listen, can you hold onto my registration and I’ll help you with what you’re working on?’. But obviously I also had something pretty significant for me to contribute in the development of Quid because of what I knew in terms of my background in science. Back at Imperial I was studying the evolution of the malaria parasites and the way you do that is you look at data for each of the different stages of the malaria parasite infection and the key part of looking at this data was looking at networks of information. So, with Quid I understood how to analyse networks from my scientific background and I also knew about how to analyse venture. And as it turns out Quid was trying to put those things together. I was in a unique position to help in tuning algorithms and figure out how users could use these insights and bring it to market.

What essentially is Quid? How does it analyse all those amounts of data?

So, the human brain is really good at understanding connections. For example, if we were to map the startups in Lisbon some of them would be fashion startups, others would be softwares for developers or computer vision companies, and so the way the brain works is it groups information together in clusters and then is able to visualise the arrangement between them. So, if I gave you a thousand pages printed out about the Lisbon companies you would manually read each one of them and you would then group together those who were very similar into piles and you would end up with sort of maps. And Quid just automates that. Quid reads information and then groups it into a map, that allows you to understand the connections between that information. If you want to read all the news about the Portuguese economy and you have a 2 day deadline for a talk you’re doing, you could have Quid do that for you and then present back the mapping of all that information for you. Quid automates the reading and the summarisation around any topic, and that automates the way humans do research.

Right now you’re starting a new company based in Lisbon and San Francisco. Can you tell us more about Muse.ai?

Yeah, sure. The company is called Muse.ai and it focusses on the problem of video. There is an increasing amount of video content on the internet and artificial intelligence can help you understand the content inside a video, like you can understand speech, vision, objects, motion, sound. So, artificial intelligence has reached the point where we can comprehend video to a similar level of comprehension of the human brain. At Muse, we’re building new ways of discovering video content by leveraging artificial intelligence. At the moment we’re focussing how to to analyse video and then later we’ll also see how to create new experiences around that. In the future you’ll be sitting in front of your TV watching a football game and you’ll say ‘show me all the goals by Messi’ and the TV will just take you straight to those highlights, through voice powered video search. And there’s always opportunities to create around how people can interact with video content.

Why build a startup in Lisbon?

My co-founder, António Roldão, is a friend of mine from Imperial College and he’s Portuguese. So, we decided to build an AI Lab in Lisbon because Lisbon has two main advantages. The first one is that the Portuguese are really good problem solvers, they are very pragmatic, and they’re open to exploration. And I think that another characteristic of the Portuguese is that they’re very honest and when you solve very hard engineering problems you need to be honest. The second advantage is that we noticed that there’s a real hunger to do something big, mainly because there’s been hard economic times but in general there’s a real optimism in doing something great. And I guess it’s also a matter of location. It’s the combination of weather, and food, and people.

And as an investor, in which industries are you particularly interested in?

I invest in a very mission oriented fashion. I’m genuinely interested 4 main types of advancement for humanity as a whole. We live in a really complex world so, can we build tools that augment our intelligence? The intersection of artificial intelligence, big data, as well as systems to sort of visualise information and be close to the human brain, like extend its capabilities, are absolutely fascinating. So, that’s the first focus. The second focus is on life extension. You know, I’m a biochemist so I have a fascination for how our bodies work and how that impacts our well-being. There are many ways in which we can build systems to extend our lifespan or sort of improve our well-being, and if we can do that we can spend more time on this planet doing better things. I’m really interested in novel mechanisms to prevent disease and ways in which we can change environmental stimulus to prevent disease in the first place. For example, what you eat has an impact on your overall lifespan and so that’s the reason for my investment in Zesty, which has gone to raise 17 million dollars. They changed what people eat at work, which is a place where people spend a lot of time. So, when you do that at scale it’s actually impacting the nutrition habits of people and that has a direct impact in their lifespan. But there are also ways of doing that indirectly. For example, I’m an investor in a company called Benchling, backed by Andreessen and Horowitz. Benchling is an extraordinary company that everyone should know about. It makes life science more efficient and it’s being used by some of the world’s top scientists as well as pharmaceutical companies. Like, there’s a great technology called Crispa, and it’s a gene editing technology so, it’s like if you wanted to turn the genes of a bird and convert it into a dinosaur. And this allows us to manipulate the genome, it can be used for a lot of negative things but also for quite positive therapeutical effects. This gene editing technology requires a lot of skill to come up with the right ways of manipulating the genome and so Benchling software helps scientists design experiments with this Crispa technology in a really efficient way. And that is really transforming the way some companies do genetic engineering as well as manage their life science experiments. I mean if you can accelerate science you can then lead to the creation of new technologies that extend the human lifespan, because it has an extraordinary net effect.

The third main theme is companies that are able to improve or accelerate digital economies. A couple of examples in that area is my investment in ClearTax, which is optimising tax return in India. Like something really great happens when you digitise the transactions that are going through an economy, it becomes more efficient. Most recently I got involved with a company called Bonsai, from Y Combinator, and Bonsai created a software to allow anyone to work anywhere in the world. So, if you’re an american company and you want to hire a Portuguese freelancer and you don’t have a Portuguese entity, you can hire them. You can hire freelancers from any part of the world. So, in the same way Transferwise makes currency conversion easier between countries, they do the same thing for labour. I’m really just interested in things that can change economies, because if you transform an economy you can create jobs and efficiencies that didn’t exist before, and as a consequence that has a net impact for humanity.

The fourth big theme is that I love companies that are building national empires. An example of a national empire is like Alibaba, a giant company that made the Chinese economy more efficient, attract the best talent in China, and then was able to create such a powerful company that they were able to go global. These national empires are companies who have grown so big within their original market that when they go global they just scare others, because they’re so big. So my goal has always been to invest in the next Alibaba. My investments in Mobius Motors in Africa, in Clear Tax in India and Chaldal in Bangladesh, are for this reason, because these companies have a chance of building a monopoly. They’re just extraordinary companies at scale.

What are your main contributions as an early stage investor? What can entrepreneurs expect from you?

Well, I actually have a methodology. I work with entrepreneurs who are in phase 0, which is typically the first 1000 decisions that a company has to make. So, within these first 1000 thousand decisions are some of the most valuable contributions you can make. The first thing is choosing an idea and making sure market timing is correct, to get the entrepreneur straight on. They also need to be the right type of entrepreneur in terms of psychological profile. They have to be irrationally ambitious and want to pursue something with deep intensity, they have to be very talented, and feel like they could walk through walls. They need to focus and be precise when they chase the opportunity,a and that’s the first thing I help with. The second thing is getting that narrative straight to help them fundraise, and typically I would join those rounds but they need to convince other investors that they’re capable of pursuing the mission they’ve articulated. And the third thing I provide value with, is that being an ambitious entrepreneur is actually a psychological game, you have to keep your head straight. You need to have courage but also be encouraged. For example, I’m a very ambitious person so I like to surround myself with ambitious people so, naturally there’s an affinity between these ambitious people, the excitement and they’re sort of motivated in high times and low times. They can always call me for that psychological council. I have all my entrepreneurs on whatsapp. If they ever need anything like an email introduction, or advice on strategy I’ll always there to help and they we all go back to work. And then the fourth thing is that I’m really good at strategy and sort of understanding how a market may unfold, like what’s the chess game you have to play to get to the end goal. My main goal in general is to help entrepreneurs understand if they’re heading in the right direction and if they’re not feeling good about their work I try to help see how I can make them feel better. And because I’m an entrepreneur myself I can be very tactical but also pragmatic. For example, looking at product and understanding if the flow is intuitive or not, or editing a pitch-deck and telling them how the narrative should be better. And once you make these things right then the entrepreneur takes over and is ready to go out. So, it’s making things that are imperfect perfect.