Thought Leaders in Artificial Intelligence: Oleg Rogynskyy, Founder CEO of People.ai (Part 2)

Sramana Mitra: What problem did you decide to go after with all this technology understanding?

Oleg Rogynskyy: We observed that most of the sales leaders at the time didn’t really know what their teams were doing. They didn’t know how to coach their sales team to better perform and have higher productivity without any data or guidance from us. It was all based on a hunch and trust. It was not based on data and knowledge.

The initial product that we did in Y Combinator in 2016 was for collecting all the activity data which included telemetry of your sales team, email, calendar, and Zoom to turn that data into a coaching recommendation of what a sales manager should be coaching their team to do.

We started with the sales team, but then we figured out that the executive has the same approach. They use the activity data to inspect what is going on and then use that to benchmark what the top performers do versus the bottom performers. They then use it to coach the bottom performers to be more like top performers. This is applicable not just in sales but also in other areas of business.

Sramana Mitra: Let’s go through a few rungs of the ladder of how you achieved what you are achieving. First and foremost, talk about data. What data are you analyzing? Give me some examples of the kinds of heuristics you are picking out of that data based on which you are modeling the AI.

Oleg Rogynskyy: First and foremost, the data we are collecting is pretty much every activity data source within your go-to-market. Your salespeople are talking to your clients via email, we collect that. This includes conversations via email, calendar, Zoom, phone, or any touchpoint that happens when a customer has a digital footprint. 

Sramana Mitra: You are collecting that off of the CRM system?

Oleg Rogynskyy: No, we collect that directly from your email system, calendar system, and Zoom system because your CRM will not have all the data. We need to have comprehensive datasets to make our conclusions. 

Sramana Mitra: So an agent is sitting on every desktop or every phone to pick up all of this data?

Oleg Rogynskyy: No, all of these activity systems have centralized APIs. We are just plugged into your centralized email fire hose API.

Sramana Mitra: Got it.

Oleg Rogynskyy: The next thing that we do, which is really important is, we classify this data into what is relevant to your business versus what is personal. This is done through machine learning. It turns out that up to 20% of the activity of the salespeople in their work email for example has nothing to do with the work they are doing. Because of that, you end up seeing doctor’s appointments, vacation planning, and communication with significant others in the inflow data. You need to filter that out.

The second thing that we do is where the real AI comes in. We need to understand why every activity is happening. For example, with this phone call between you and me, questions will pop up like, “Is Sramana recruiting or is it the other way around? Is Sramana investing in People.ai? Is Sramana interviewing? Is Sramana looking to buy People.ai products?”

It’s about understanding the context of every communication and the reason why it happens. This is required for us to attach this communication to the right place and the right system of records such as an opportunity in your salesforce record or a candidate in your ATS system. We have built an AI system that understands the nature of every activity and why it happens so that it can be attached to the right place. 

This segment is part 2 in the series : Thought Leaders in Artificial Intelligence: Oleg Rogynskyy, Founder CEO of People.ai
1 2 3 4 5 6