Thought Leaders in Artificial Intelligence: Salil and Kiran Pande, Co-Founders of VMock (Part 5)

Sramana Mitra: Between 2009 and the time you raised your seed financing, you figured that piece out. You figured out that you were going to go after the US career development offices and you were going to build this product in India.

Kiran Pande: Yes, he was leaning towards having the engineering team from people in IIT. The Bay Area is also very expensive and competitive to hire from. Just the cost of building an engineering team in the Bay Area would have been a lot more than we were able to do at that time. 

Sramana Mitra: Did you have any of the products done at the point you raised seed money?

Salil Pande: Seed money was very simple. It was concept-driven.

Sramana Mitra: People were willing to write a check to let you build a product to sell to career development offices while you were doing the product development in India. This is 2012?

Salil Pande: Correct.

Sramana Mitra: Between seed and Series A, how did you bring together a version one product to deliver something for which career development offices are willing to start paying for? Series A funding doesn’t happen without paying customers. What happened in those three years? What did you build? How did you train the models?

Salil Pande: By then, we had figured out that interviews are going to happen, but it has to wait. We slowed down on Interview or any of the experiments that we were doing. Resume stood out as a product that we needed to focus on. We built our initial resume product and then we launched it to several institutions in the US.

Once we had a pretty strong proof of concept, that’s when we raised a Series A. It was also clear that it’s not just going to be a resume product; it’s going to be a whole portfolio of services and products. We used to call it smart career platform. Later on, we decided on calling it a career acceleration platform.

Sramana Mitra: But the first proof of concept was on the resume product?

Salil Pande: Yes.

Sramana Mitra: Can you elaborate on how you train the models for the resume product?

Salil Pande: We’ve blind data. Every institution that we work with, we have some historical resumes.

Sramana Mitra: Let me probe here. Let’s take an example. Let’s take MIT. Let’s say you are working with the MIT career development office. We have a lot of departments. There are resumes with various kinds of skillsets and degrees. In terms of resume development, give me a couple of examples of what you can do that lets you work across all these different departments.

Salil Pande: If a freshman and a senior are working on their resume and trying to get feedback, it can’t be the same. We make sure that our feedback is commensurate. If MIT calls its business school as MIT Sloan, then it should be properly represented as MIT Sloan. If they have a format that they actually adhere to on their student resume, that is the format that students are able to align with. 

Sramana Mitra: There’s nothing domain-specific. Is it all formatting?

Salil Pande: There’s a lot of domain specificity as well. If you look at the way we represent ourselves, we have built our machine learning parser. We have built a job description parser. We have built a skills engine that has more than 25,000 skills that all get an action to provide this feedback. 

Sramana Mitra: One of the constraints that I noticed in your business is you’re working with career development offices, which means that the bulk of users are students. They don’t have 10-year resumes. They have very short resumes with mostly academic work. Grad work and internship potentially. That’s one constraint I’m observing.

Kiran Pande: Yes and no. You have to think about the journey. We started out with business schools.

Sramana Mitra: That’s news to me. When you started, you started with just business schools.

Kiran Pande: Now everyone can use the product. There’s no limitation on numbers of years of experience.

Sramana Mitra: By the decision of selling to career development, you’re selling to people in schools right?

Kiran Pande: Yes. Now we’re no longer in schools. In the business school segment, we made sure that we perfected our product. We went to other segments and large universities. When we started out, we knew we wanted to build something that works for everybody. At the same time, we had to start somewhere and perfect that, then expand from that.

This segment is part 5 in the series : Thought Leaders in Artificial Intelligence: Salil and Kiran Pande, Co-Founders of VMock
1 2 3 4 5 6