Thought Leaders in Artificial Intelligence: Zohar Bronfman, CEO of Pecan.ai (Part 3)

Zohar Bronfman: Lastly, you also want to compute the label. The label basically means whether this customer has churned 30 days later. To generate the label, the platform will automatically take this transaction and identify whether there has been a churn event.

Let’s say the user has defined churn as 30 consecutive days without any transactions. The algorithm will look at the timestamps of the transaction and generate a label of one if the customer meets the criteria, and zero otherwise. With those business definitions, we get in the UI along with the information about how the tables are related to each other. The platform is able to generate the whole AI-ready structure that consists of those variables that I mentioned earlier. 

Sramana Mitra: What are you seeing in the market as competition?

Zohar Bronfman: We are best-suited for cases where the potential customers don’t have a full-blown data science capability but have strong analysts. For those types of organizations, we help them adopt data science to their existing analytical resources. The real type of competition for us would be data scientists. The organizations that consider using Pecan are organizations that are also considering hiring data scientists that would build those models. 

Sramana Mitra: It sounds like churn management is one major use case. What are the other major use cases?

Zohar Bronfman: We have the likelihood for a conversion or the likelihood for an event. You can predict whether a customer is going to make a deposit or going to make a subscription. You can predict the lifetime value of an individual. You can predict the likelihood for upsell and next-best offer. On top of those, the last use case is demand forecasting, which is an answer to the question, “How many transactions are going to be made in an aggregated manner for a specific product?” 

Sramana Mitra: What industry segments are you seeing the maximum activity?

Zohar Bronfman: Direct to consumer or consumer services companies. Also mobile applications for entertainment, media, fintech, social gaming, and retail.

Sramana Mitra: Telecom perhaps.

Zohar Bronfman: Yes, we also have telecom companies.

Sramana Mitra: At what point do you plan to open this platform up for other developers to build applications on?

Zohar Bronfman: This is part of our strategy. As we look at our roadmap, it will probably be early 2022. This is still in development and testing.

Sramana Mitra: When you open up your platform, what areas do you foresee as white spaces?

Zohar Bronfman: Let’s define the kind of developers that we are counting on. Those would be our users. These would be BI and analysts that can work the SQL, analyze the tables, and generate classic analytics insights. Those users will be able to improve on existing templates that I mentioned. Building additional templates would also be useful.

Sramana Mitra: What I’m asking is more of the different kinds of use cases where you can broaden your scope.

Zohar Bronfman: Two come to mind. One is questions that are second-order. Not only what would be the demand but what would be the optimal pricing. It’s a second layer. The second is the field of chargebacks.

Sramana Mitra: Very interesting. It was a great discussion. Thank you for your time.

This segment is part 3 in the series : Thought Leaders in Artificial Intelligence: Zohar Bronfman, CEO of Pecan.ai
1 2 3