548th 1Mby1M Entrepreneurship Podcast with Lloyed Lobo, Co-Founder of Boast.ai (Part 4)

Sramana Mitra: Let me probe one thing here. When you were doing this manual customer servicing, was it more of a rule-based engine than pure AI?

Lloyed Lobo: First, it was fully manual. The rules-based engine was the Wizard-of-Oz MVP. It looks like technology but, on the backend, humans are doing it. There are two elements. You are collecting data. Then you need to process that data and generate the outcome. Then we started doing this data collection online where people can upload everything.

You need to normalize data. Data that is coming from payroll or bookkeeping is already in ones and zeros. Data coming from your GIRA and GitHub is unstructured data. Trying to figure out what projects qualify and who spends how much time on it is the hardest part. That is an ongoing project for us to make it better and better. The third piece is how do you look at these data points and then automatically write reports for the government. It’s like natural language generation. 

Sramana Mitra: But your bootstrapping phase was manually delivering how much tax credit would a company qualify for and the application to the government. That was your deliverable, right?

Lloyed Lobo: Exactly. For the client, I got a check from the government and I did not get a government audit.

Sramana Mitra: Interesting. You have explained in the beginning that your business model is that you’re financing that tax credit. The business model was different in the beginning?

Lloyed Lobo: We have two business models. Both models are, more or less, the same. We take a percentage of the tax credit after they get it. The financing just helps us bring forward the credit. In November, we closed our Series A. We did a $23 million Series A. In February, we did a $100 million warehouse credit facility to lend against those tax credits. Entrepreneurs sit on these tax credits for a year, file it, and then get it. The financing is a new thing.

Sramana Mitra: Initially you were just providing the service of figuring out how much tax credit people would get and the application generation for that tax credit. What was the business model before? Paying you a percentage of the tax credit that they would get back?

Lloyed Lobo: Exactly.

Sramana Mitra: Before or after they got the money?

Lloyed Lobo: After the check was in their bank.

Sramana Mitra: It’s a very long cycle though, right? Governments take a long time to process this.

Lloyed Lobo: Yes, a very long and stressful cycle. What Avalara did for sales tax, we’re doing for R&D tax credits. The big four predominantly do R&D tax credits. They charge you by the hour. We wanted to break that mold. I call that value-based pricing where they get value and pay us. It was a very long sales cycle. It was so stressful. Now we’re at a point where we have so many customers. Although the cash cycle is long, we’re cash rich. 

This segment is part 4 in the series : 548th 1Mby1M Entrepreneurship Podcast with Lloyed Lobo, Co-Founder of Boast.ai
1 2 3 4 5 6