Bootstrapping First, then Raising Money to Build a $10M+ Generative AI Startup: Anthony Scodary, Co-Founder of Gridspace (Part 6)

Sramana Mitra: Let’s go back to your company building. You bootstrapped for a year and a half to product. Then you had started getting these paid pilots. Did you raise any money or is it fully bootstrapped all along?

Anthony Scodary: No. We mostly grew off of revenue, but we mostly took money from strategic investors like Stanford Hospital invested in us after we did StartX. Our investors include some of our customers like USAA and Bloomberg. Wells Fargo and Santander also invested in us. As we got bigger, through friends of friends, we met investors who were a good fit for us. It gave us the opportunity to build bigger models earlier and build the product out a little bit faster. But this was after the point where we had found a market and we had found folks that relied on our tech to do their job.

Sramana Mitra: So how much money have you raised outside capital?

Anthony Scodary: It was over $7M.

Sramana Mitra: Okay. Where are you today in terms of metrics that you feel comfortable divulging?

Anthony Scodary: Our ARR is in the eight figure range. In general, it’s all folks that are mostly using the voice tech for one or two use cases. One is the foundational analytics model for our largest partners. The second use case is automation, either on the passive side where we’re automating parts of a human call center or they’re rolling out Grace.

The majority of our revenue comes from SIFT, our analytics product where we’re passively analyzing. We attach to your call center, analyze all the voice in your call center live, and then we provide analytics and the ability to get that to data scientists, get that to people who are doing call center ops. Our biggest growth area is the virtual voice bots.

Sramana Mitra: Okay. Is there anything else that you want to share? Are you based in the Bay Area, by the way?

Anthony Scodary: We’re based in both LA and San Francisco. Rright now I’m in LA, this is our main office. We have some folks up in the Bay.

Sramana Mitra: What about the rest of the team? How big is the team and how is it distributed?

Anthony Scodary: It’s thirty people and the majority are in Los Angeles. The team is mostly, engineers, scientists, and designers.

Sramana Mitra: Apropos our earlier conversation, what you were saying about using your tools to provide solutions, all this analytics work that you do, you actually do the solutions. Thirty people are enough to do what you’re doing and charging a lot for. It sounds like a highly profitable model.

Anthony Scodary: It’s challenging, because we’re competing with companies that have thousands or tens of thousands of employees or more. In the startup world, if you have very good focus, you have very good product fit in general. We are mostly engineers, because we’re trying to build state of the art machine learning models. In general, it doesn’t take a giant army to build a good product.

Sramana Mitra: No, it doesn’t take a giant army to build a good product. If these engineers know how to use the product and how to build the solutions on top of the product , that also doesn’t take that long if you have the right people with the right knowledge.

It’s not ten thousand people who are going to have the deep knowledge of this product. What you’re able to do is to learn about the way AI solution deployments are happening. It’s with highly capable engineers who know how to build platforms, tools, and technologies and then build solutions on top of them.

You can really build very, very profitable companies with that combination. And that’s what a very good case study are.

Anthony Scodary: Yes, I completely agree. I just got back from recruiting at Stanford yesterday. Almost the whole team came out of Stanford, MIT, and CMU where there’re top speech programs and research. If you go to places, even like Open AI, mostly people would just want to do research. They want to make papers, not products.

When you’re at a product-focused startup and trying to build something and get it to the market, you need people who want to build whatever needs to be built. We believe in longitudinal ownership where the same person is collecting the data, building the model, the backend, the front end, and doing the operations and the deployments.

The reason that’s really important is the Conway’s law. It basically says that when you build an engineering system, it tends to have the same form as the team that built it. So the original example in Conway’s law was a compiler where, the number of passes in your compiler tends to be almost always the number of people who worked on the compiler.

When you have teams where you have artificially segregated roles, you have all these extra abstractions and interfaces and tend to not have a lot of ownership. When you have engineering generalists who know machine learning and backend and front end development and who are just focused on building the best product, you can get not only a lot more done, but you also have a much cleaner product with a lot less complexity. It’s a lot more robust. Also, if someone asks about a component, the cognizant engineer can answer a question about any of it, not just their little piece.

Larger companies have the luxury of having very inefficient scale because of large teams with lots of segmented roles. At a startup, you generally don’t have that role. It’s a cliche that people wear lots of hats, but in engineering, what that really means is you can’t just be particular about the type of the research you want to do. We’re not trying to publish papers. We’re trying to build product.

This segment is part 6 in the series : Bootstrapping First, then Raising Money to Build a $10M+ Generative AI Startup: Anthony Scodary, Co-Founder of Gridspace
1 2 3 4 5 6 7