Mint brings together three entrepreneurs who represent the new dawn. Excerpts from a conversation with Derick Jose, co-founder of IoT startup Flutura, Aruna Schwarz, co-founder of AI company Stelae, and Balaji Viswanathan, co-founder of robotics startup Invento.
How did you become an entrepreneur?
Aruna Schwarz: I got my first flavour of entrepreneurship in a joint venture between Cable & Wireless and Schlumberger with a carte blanche to build an internet product for the oil and gas sector in the late 1990s. I was living in Paris after doing an MBA in the UK. I exited telecom to get into software with a small funded company in Paris, transforming PDFs into publishable outputs.
That was in the early 2000s when everybody wanted to put content online. The problem was turning all these documents into HTML. There were a lot of cross-media publishing systems but I wanted to build something that wouldn’t need a change in human behaviour.
I met my CTO Pierre Fraisse who had a PhD in maths, spent over 20 years in R&D on content management, and grown in the publishing sector in France. I didn’t know code but what I did know was how to take a product to the market. We started building our product for a large French publishing company, Lagardere. We later moved into legal publishing because there was no money in online media.
A lot of people think startups are there to change the world. Maybe, but not all of us, because you need to make money to sustain the people who work with you.
How did you get your first big break?
Schwarz: It was just Pierre and me until 2010 when Rolls-Royce in the UK came to us and asked whether we could apply our algorithms to their documents. They purchased a six-figure licence from us and that’s how we got into aerospace.
Then I moved the company to India in 2012 when we got a huge digital transformation project for defence. We’ve worked with the French naval defence company DCNS and last year we signed a contract with Airbus. Now we’re purely focused on aerospace and defence.
We expanded with customer revenues and we’re now a team of five. All the algorithms, the way the software is architected is Pierre’s work. It’s not a short-term game because when you really start to scale is in the seven-eight-year period. I don’t know what Derick and Balaji would say, but look at SAP or Oracle. Sustainable products and companies take time to build—without becoming unicorns.
Balaji, how did you become an entrepreneur?
Balaji Viswanathan: I grew up in a small village near Kumbakonam in Tamil Nadu. My father was a bank manager and he used to take me to all these places where he would give loans to entrepreneurs. It fascinated me to see these people make something out of nothing. Somebody was doing mushroom farming; another one was making scientific test tubes and beakers.
In my own family, my grandfather quit his bank job to start a company making agar-agar from algae. That went down because it wasn’t managed properly. But my maternal grandfather came from Burma during the World War with nothing, started a small shop and grew that into a printing press. So I saw both the negative and positive sides of entrepreneurship.
I started my career at Microsoft in Redmond. After a year of building Windows, I was put in something called Startup Factory. This was around 2006 when Google and other startups were building all sorts of cool things. We were taken far from Redmond to be out of its influence and told to build things like they do in Silicon Valley. Small teams got short-term funding and after six months you had to raise another round like in a VC model. It was an interesting experiment but then the recession hit in 2008 and I was asked to help with Windows Phone. We did not like being moved like that, so most of us quit.
My partner and I came back to India and started an edtech company called NalandaU. We got content from top universities and built a social layer around it. This was 10 years ago, before Coursera, Udacity and others. But it was hard to convince VCs that Harvard and MIT would give their videos for free and people would watch these videos. Then we pivoted to fintech. But this was 2011 and VCs said the financial sector was the slowest to adopt technology, so why would we even want to call it fintech.
We wanted to extract sentiments from social media and see what the smart money was up to. The same guys who said that was the stupidest idea they had heard would be funding the same thing with another company three years later. So what I learned from all this is that you have to survive the first three to five years to make a mark.
So Invento Robotics is your third try...
Viswanathan: After a stint with a Boston company helping enterprises manage open source software, I moved back to Bengaluru to start Invento three and a half years ago. This time we decided we would not even talk to investors for the first three years, because if we listen to them, we might take them seriously. We were designing and building hardware with AI on top of it. This is something everybody will say the Chinese can do bigger, faster, better. So you have to shut your ears and stay on course. Only after crossing the three-year mark, we raised two rounds of funding.
Derick, you moved out of IT services to become a deep tech entrepreneur. What made you do that?
Derick Jose: Three of us—Krishnan Raman, Srikanth Muralidhara and myself—were cruising along at Mindtree, doing templated stuff. But one day this question crept in: Are you an order-taker or an order-maker? Suddenly it didn’t seem so cool being order-takers. We wanted to do more order-maker stuff.
The three of us started going to our existing customers and asking, ‘What if’. Each time, the ‘what if’ was shot down. We didn’t know why. So on one of our travels abroad, we sat down over a coffee with a manager and asked, ‘Why wouldn’t you want to do it?’ He said, ‘These ideas are great, but you know what, we’re coming to you because you guys are cheap, not faster or better.’ That was a significant inflection point. That’s when we had the intoxicating idea that we would create something faster and better that will not be cheap.
For two years, that keeda (worm) kept going around our heads. We couldn’t do it within the existing structure because it was oriented towards people who only perceived value in terms of cost. Then Krishnan went and announced that we three were quitting. Srikanth and I came to know we were quitting only after he announced it.
How did that work out?
Jose: The first six months were all about the excitement of being free in a cool new world, brainstorming. Then we had to start converting. The first question was: Will somebody write a cheque for us? We used to get orders at Mindtree but we didn’t know if it was for the brand or what we did. We managed to get a couple of orders from Cleartrip and a London healthcare company. It gave us the confidence that we could sell something as individuals and deliver it.
Next, we had to build a product because we didn’t want to do services. But this meant going without revenue for a long period of time. Our savings were not enough to do that. Those days we used to write a lot of blogs on the big data theme. One blog caught the eye of Arihant Patni (whose family built and sold Patni Computer Systems). He came to our office. I don’t know what we said but the three of us were talking a lot. He saw the energy and decided to write a cheque. That was our second moment of truth.
Our vision was to be among the top three in our category of AI applications for oil and gas and speciality chemicals. Last year, we were third in Gartner peer insights, which ranks vendors based on customer ratings. This year we’re number one, beating Siemens, Hitachi and others.
At this juncture, do you see any opportunity for cross-learning between the three of you?
Viswanathan: I can learn a bunch of things about walking into unfamiliar spaces. How do you bring clients to the table?
Schwarz: My experience there is that we’re into the hardest cartelized industry which is aerospace and defence. For me, the way of getting around that was to have a French CTO. And I do have 35 investors, many of them in Europe, all private guys. That gave me leverage. And we’ve been around for over seven years, so they know we’re not going to die.
Breaking into one industry and staying there is important. We went to one of the hardest but we got there because we got a call from Rolls-Royce. Another lesson that Derick and I will agree is that it’s not always about the best product but who you know and who knows you, which involves a lot of hard work in being at the right place at the right time.
Jose: I look at all three of us as seeds. If the seed falls in Bengaluru, then the soil here kicks in. We can say Flutura is God’s gift to mankind, but it will not take us far unless we can also brand the soil around us. That’s why we’re obsessed with branding Bengaluru and India. Take Israel: the worst Israeli security startup will beat the best Indian security startup in perception because of the soil factor.
Jose: The perception about the country is vital. When we send a rover to the moon, we should not be saying it cost us less than making a Hollywood movie.
Schwarz: I know, that is so embarrassing.
Jose: When a person is buying from us for industrial operations, he’s not looking for the cheapest solution. He’s saying I’m ready to pay three times the price. So when we talk of Isro (Indian Space Research Organization), we should not be saying we’re cheaper. We should be saying we’re faster or better.
Viswanathan: Or more reliable.
Jose: We’ve got into this psyche of cost, cost, cost. So they look at us and think we’re cheap. That’s where we want to see a shift at a country level. We won’t be cheap. We’ll be at the same price.
Schwarz: I hate that word ‘cheap’, and I think it’s useless because nobody wants something that’s going to break. But this is what the Indian coding industry is about. When you go into some of the European manufacturing companies, they don’t think Indian software is reliable or sustainable.
They don’t think software products coming out of India are real products either. They think it’s a cobbled-together application we built for somebody else and are trying to resell. So building the India brand for hardcore software products has to be done and that’s why Derick and others have pioneered the industrial AI forum.
Jose: We need a new narrative where we’re able to say we’re the world’s first at something, even if it’s in niches.
We may not have them in consumer AI but we do have some verifiable firsts in subcategories of industrial AI. Take a startup like Detect which can easily say they’re the world’s first company in AI inspection of oil pipelines. Mitra robots could be the first in XYZ subcategory. We can say Flutura is the world’s first in autonomous machine data gathering systems.
We’re taught to be humble in India, so we face severe resistance in being able to say we’re the world’s first, even if we’re actually doing it. I got important feedback from an investor, T.M. Ravi, in the Valley. After my pitch, I asked him what he thought. And he said, ‘Derick, you should walk like a samurai.’ We’re ‘samuraizing’ our product, why can’t we ‘samuraize’ our body language?
Viswanathan: A related point is that many of the new startups don’t project themselves as Indian startups. For a while we were following a startup with facial recognition technology that we wanted to integrate with ours. They called themselves a Singaporean company and we were talking to them on Skype.
At some point, we asked where they were sitting and it turned out to be four buildings away. I said, ‘Dude, I’ll switch off Skype and walk to your building right away.’
So the problem is that even successful brands coming out of India project themselves as an American company or a Singaporean company because our country does not have a brand. It becomes a self-fulfilling prophecy because these companies won’t be known for their Indianness. That’s not the case with Israeli startups.
Schwarz: Totally. There is an Israeli brand which is reinforced by the bigger Israeli brand in the US or Europe. So, echoing what Derick said, we’re known for being cheaper. That’s the whole service industry.
And even in the software product industry, the only ones making news are SaaS products which are cheaper versions of something else. Where we should play on is a hardcore differentiation in our product. Our software, for example, is the only one that does automated conversion of aerospace documents.
Jose: World’s first.
Schwarz: Yes, I can say we’re the world’s first.
Jose: There you go!
Schwarz: For me, it’s easier because being a woman, you know, they don’t expect you to say that. So when I do say it, they think it’s highly cute, whether they believe me or not.
Viswanathan: At least Europe and the US have been doing business with India. East Asia, especially China, has a far lower opinion of India. We’re also not doing favours to ourselves. They know Indian merchants will go for the cheapest things.
But China is also becoming more of a consumer society than a producer. The cost of doing business in China is rapidly increasing. We’ve been talking to multiple factories there. In Shenzhen, the cost of entry-level factory labour is ₹80,000 a month which is four or five times that of India. The cost of land has shot up in the last 15 years, and cost of regulations also because now they seriously care about pollution. A couple of factories asked us how to set up bases in India. Textiles have moved out of China to Bangladesh and Vietnam. Other industries have moved to the Philippines. This is actually a time when we should be entering China with our products, but again it should not be on the cheap selling point but quality.
How are you making your China play?
Viswanathan: So one of the tags we’re using is that we’re one of the first enterprise-ready robotics companies. There are a bunch of robotics companies who can move hands, fingers, all those things.
But what we say is that you can connect your enterprise software, CRM, email, etc, to our robot’s software. That is something they start to believe because there are so many enterprise software companies in India. So our robotics software is ready to get integrated with enterprise software. That is the hardware-plus-software story that is starting to resonate for us in China.
I’ve been travelling to China every three months for the last couple of years. We’re now incorporating in Suzhou near Shanghai. One thing about China is that it is very provincial. There’s a lot of competition between municipalities. So even if it’s a Chinese company, they will say, ‘You’ve not incorporated in our municipality.’ So this creates an opportunity for us.
Does it bother you that deep tech or hardcore enterprise tech startups get so little funding in India compared to the consumer internet startups, which are raising millions or even billion-dollar rounds?
Viswanathan: A couple of startups on the next street from us in HSR Layout became unicorns in just two or three years. One of them became a unicorn before having a viable product, so that’s pretty crazy.
Schwarz: Yeah, I know. It became a unicorn on valuation. But if we get depressed every time somebody gets funded, we would have chronic depression.
I do get annoyed when me-too models get millions in funding. But the thing is, for me, I get a bigger buzz from a purchase order than I get from investor money.
Viswanathan: One of the companies in our neighbourhood is three-four years old and they notched up losses of ₹52 crore with revenue of ₹10 crore. How do you do that? I see them as being in a different world from ours. I keep working with my CFO on how to get to profitability fast, although VCs tell me that’s the wrong thing to focus on. But fundamentally, I feel more secure if the company is actually making a profit.
Jose: Most of our global rivals are also much better funded than us. But when you’re an underdog, I think it makes us more capable in some areas to win David versus Goliath battles. Some of our rivals with a lot of money have become a little bit comfortable. So that’s a positive for us.
Viswanathan: Also, to see companies grow so fast around us is invigorating. Four buildings from us, there’s a company building control systems for electric scooters in China. This team goes there, deploys, and comes back to HSR Layout.
I’ve lived in San Mateo, Palo Alto, Seattle and Boston. But here, it’s an incredible experience to walk across and see the hardware and software that startups are building in all kinds of areas like aerospace and so on. It’s magical.
Ten years from now, we’ll be looking back and saying we underestimated what was happening. Like you guys, Derick, are selling to oil and gas majors in different continents.
Schwarz: For me the biggest change since 2017 is that the US and European giants are picking companies from here for their innovation labs.
The fact that they’re nurturing technologies, whether it’s in aerospace or retail, and taking them global inside their own organisations is the phenomenal difference I see because they’re actually buying products from startups.
And many of these startups I had never heard of earlier, and I don’t think Nasscom had heard of them either.
Jose: India is a great place to cross-pollinate, unlike other clusters. Houston does not have AI skills. Bay Area does not have electromechanical skills. Here in Bengaluru, we have both electromechanical and digital skills, the old and new worlds constantly colliding.
To wind up, what are some of the funniest experiences you’ve had as entrepreneurs?
Schwarz: When I used to go to VC meetings with my co-founder Pierre Fraisse, they would call him “Mr Schwarz". They just ignored me and talked directly to him. At first, I used to be upset but then I was like, this is so funny. They say the craziest things and I can sit in a corner and watch as an observer.
Viswanathan: We were showcasing our humanoid robots at a tech event in Dubai. We allow a range of expression of fingers and gestures to have a more human connection. But an API call went to the wrong digit and our robot showed the middle finger to a senior government official instead of a thumbs-up. At first, the official was taken aback because he had his subordinates around him, but then he started laughing and gave the robot a playful slap.
Balaji Viswanathan is CEO and co-founder of Invento (founded 2016), whose humanoid robot Mitra greeted Prime Minister Narendra Modi and US President Donald Trump’s daughter Ivanka on stage at an event in Hyderabad. These days, Viswanathan is a frequent visitor to China where his robots are in a race with their Chinese and Japanese counterparts, hoping to outflank them with their AI brains and enterprise smarts.
Aruna Schwarz is CEO and co-founder of Stelae (founded 2002), whose AI product Khemeia turns legacy documents into structured content that can be integrated with digitized systems. It does in minutes what would take days to do manually. From legal documents to aircraft manuals, Khemeia plays a part behind the scenes in the digital transformation underway in multiple sectors. Schwarz’s fluency in French opens doors for her in Europe.
Derick Jose is co-founder and chief product officer of Flutura (founded 2012), whose industrial IoT product Cerebra uses AI for predictive maintenance and efficient processes. From oilfields in Houston to German multinational Henkel’s plant near Shanghai, Cerebra is in mega industries, taking on global rivals. In his spare time, Jose is building an industrial AI consortium to help Indian startups overcome perception hurdles abroad.