From Developer to Successful Machine Learning Entrepreneur: Aparna Dhinakaran, Co-Founder, CPO of Arize (Part 5)

Sramana Mitra: So what use cases did you find the maximum traction in?

Aparna Dhinakaran: AdTech click-through rate is a common use case. Fraud is a big one. Lending is a big one. Demand forecasting is another big one. There are these common use cases.

Sramana Mitra: Are you direct selling?

Aparna Dhinakaran: Yes, direct to ML engineers. We pitch and show them the product. Engineers typically just want to see the product.

Sramana Mitra: You had the product at a level where an engineer can take it and run it on their own?

Aparna Dhinakaran: Yes.

Sramana Mitra: How about pricing?

Aparna Dhinakaran: At this stage, what we’re doing is close to releasing a self-serve option. Today, you talk to sales and there are pricing structures around the volume of predictions. As you move towards more of the modern sales approach, it’s using volume-based pricing that’s very transparent on the website. It’s usage-based. That’s what we’re transitioning towards.

If there’s one takeaway from the space, the big one would be that this space is going to be massive. You think about Datadog and Grafana. When they started, people were probably asking why you need observability.

Sramana Mitra: When you think about TAM, you probably have a number in mind in terms of how many machine learning engineers and models there are. How many teams are actually adopting machine learning observability solutions at this point?

Aparna Dhinakaran: That’s a great question. It’s very early. It’s probably less than 10%.

Sramana Mitra: That’s my estimate as well. I don’t think this has started getting critical mass.

Aparna Dhinakaran: Exactly. The thing you get asked is how big can this be. You want to pick something with a big market. Cloud spend last year was somewhere north of $200 billion.

Sramana Mitra: That’s too far removed from your specific TAM.

Aparna Dhinakaran: I think AI spend was around a tenth of it. The reason why the infrastructure market is a good comparison is that the infra market is around $200 billion. It supports all of these observability companies. The market cap of Datadog and Sumo Logic is $70 billion, so a third of the $200 billion infra spend is the market cap for infra observability.

If you think something similar or parallel is going to happen in the AI spend, even if I take a tenth or a third, that gives you the vision of how big the market can be.

Sramana Mitra: I have no doubt that this is going to be a big market. Talk to me about your entrepreneurial journey as a woman entrepreneur. Do you believe there is a bias? Have you experienced any pushback?

Aparna Dhinakaran: It’s an even unbiased journey. We don’t teach, as a society, that it’s okay to fail. It’s 100% okay to fail. We especially don’t teach women that it’s okay to fail. A big part of this is, you got to fail a thousand times and get back up. You just don’t give up.

Sramana Mitra: All of those are points well-taken. You have been in the Bay Area where the failure culture is well understood. You come from a Berkeley CS background and you’ve worked at Uber. Your resume is very strong. If I am an investor looking at you, the fact that you’re a woman is not going to be a factor for me. I would argue that we are at that point. In Silicon Valley, at least.

Aparna Dhinakaran: I don’t see it in reality though. Why aren’t there more women AI founders?

Sramana Mitra: They have to have that resume. There are a lot of factors. There are not enough women AI founders because there aren’t enough engineers.

Aparna Dhinakaran: It’s not just raising money that is hard. You got to go on every single sales call and pitch. Why should they buy a product from you versus some of your male competitors? It may be implicit.

This segment is part 5 in the series : From Developer to Successful Machine Learning Entrepreneur: Aparna Dhinakaran, Co-Founder, CPO of Arize
1 2 3 4 5 6 7