1Mby1M Virtual Accelerator Investor Forum: With Ankit Jain of Gradient Ventures (Part 6)

Sramana Mitra: It’s fascinating, isn’t it? It’s a really exciting field. Are you chasing unicorns?

Ankit Jain: Given that we’re investing $1 million to $10 million, we’re looking at a lot of companies that we hope will show their true market potential. We hope some of them get to that stage. We strongly believe many of them will get there.

Sramana Mitra: AI, as you said, is applicable to every domain. If you understand the tools of AI, you can apply it in creative ways to solve problems in every single vertical. Not all of these are billion-dollar TAM opportunities. Some of them are specialized. How do you parse these opportunities?

Ankit Jain: We’re still at a very early stage in figuring out the true impact of AI on many markets. Early stage investing is understanding the potential impact of a product. Market sizing evolves as markets expand. We’re still way too early to understand the true market size in many of these fields.

What we look for are the right founders in the right market. For an early stage investor, that’s what you should be focusing on. If companies do things the right way, market sizes will expand. The best examples of this is Uber. If you look at the global taxi market size, people would not have been able to imagine how big Uber and the entire ride-sharing ecosystem is today. The market has expanded with the success of the Ubers and the Olas of the world.

Sramana Mitra: Yes, true. Going back to your question about founder-market fit. One of the requirements of being able to define really interesting AI solutions to domain-specific problems is deep domain knowledge. I would say in evaluating founder-market fit, you need deep domain knowledge in a particular workflow and in a particular domain to be able to define a good solution to a vertical problem.

One of the things that’s happening on the large company side is, there’s an abstraction happening where the software layer is being tackled at the platform level. There are provisions now for people with not rocket science level knowledge of those deeper layers to be able to define solutions specifically with this idea in mind.

If you have a founder who has maybe created a partnership with Google’s AI platform and has built something on top of that using his or her deep domain knowledge, you have a great founder-market fit for that. However, it’s a very niche solution. It’s not a billion-dollar opportunity. It’s a hundred million opportunity. There are lots of such opportunities out there. This founder’s domain knowledge is in that domain. Expanding out of that doesn’t give them a leverage.

Ankit Jain: If I may, I’ll go back to my previous example. That was a very niche opportunity when it started off, but they created a large market within that niche. Then you have the Oracles and SAPs that have designed the database systems for many large companies. You could argue that systems built on top of that wouldn’t have the ability to be large companies. But there’re companies that use database. They probably do not develop their own. Workday, for example, is a very niche HR opportunity, but they’ve made a big business out of it.

Sramana Mitra: HR is a big horizontal right? What you’re describing are big horizontal opportunities. What I’m talking about is the more specialized verticals.

Ankit Jain: Sure. Fundamentally it’s hard to talk hypothetically. I’m sure there’s going to be opportunities where we see a good business, but not one that’s a good fit for venture. That’s perfectly okay. Like a lot of funds, we work within the constraints of what a venture model is. We’ve got our internal benchmarks on how much growth we’d like to see for it to make sense.

Sramana Mitra: I got that. This is not applicable to your fund. I’m just pointing out that there is a lot of this stuff going on. I talked to the CTO of Microsoft Azure where they’re really trying to provide that abstraction layer so that people who want to use AI within a specific domain can avail of the AI facilities.

I think it makes a lot of sense because the rocket scientists are going to work on startups like the ones you are investing in. They’re not going to be working on these niche vertical software in some corner of automotive.

Ankit Jain: This is the power of platforms whether it’s platforms like Microsoft Azure or Watson. I think Wix is a great example that enables a lot of folks to make websites because they said, “All you need to know is an idea of what you want it to look like.” If you want to have a wedding website, Wix is probably one of the best solutions for it.

Sramana Mitra: Exactly. It’s a bit more complicated to do that with AI, but that’s really where it’s going.

Ankit Jain: It’s complicated today. If we were to look at this in five or seven years, it wouldn’t be as complicated, just like when you’re creating a site on Wix. On the background, it’s using databases and APIs the same way. I think there will be layers of tools and layers of abstractions that will make AI be part of it without you knowing it. We’re noticing the first flavors of this.

There’s a company in our portfolio called Cogniac that takes the magic of image labeling and building a model around it. It abstracts it out so that some of its customers are feeding them a bunch of image data. They figure out which are the images that need to be labeled in order to build a model with high accuracy.

There’s other similar work going on with Google. The idea here is to enable as many folks and as many industries whether niche or large to not have to have a deep understanding of AI. Just have a deep understanding of the domain and then go build something on top of it.

Sramana Mitra: Fascinating. It was a great conversation. Thank you.

This segment is part 6 in the series : 1Mby1M Virtual Accelerator Investor Forum: With Ankit Jain of Gradient Ventures
1 2 3 4 5 6