Thought Leaders in Big Data: Sanjay Vyas, CEO of Diyotta (Part 2)

Sramana Mitra: Can you talk about a few use cases? I would like to do three use cases. Double-click down on what exactly you are doing.

Sanjay Vyas: Let’s take the first use case as customer churn, which is a well-identified problem in the telecom industry. When you are a large telecom, you are trying to retain your customers, but you also want to know what behaviors and patterns are there that indicate if they are about to leave you.

That churn prediction is very tough if you are just reacting to the customers or looking at it as a post-mortem. To be proactive and to see whether there are any patterns of the customers leaving, they will need to anticipate those patterns. Is the customer trying to interact with them on support channels, social media channels, or on the telephone IP? Are they getting frustrated? Are there any repeated behaviors there? If they are not getting the customer satisfaction that they deserve, how will you bring all of those data points together and consolidate them, how will you collect that data from all these different channels?

That itself is a big problem right now where you need to marry your systems together and get the data in time so that you can predict the churn. Let’s say the customer came from the support channel. Then they go to social media because they didn’t get the answer. You are then able to understand and say, “Oh, it’s the same customer that had some issues in the support channel. Let me have a look at it and see if we can resolve it.”

That is one of the problems that we were able to help large telecoms with. We provided proactive data analytics to reduce the churn in their customer retention. That is one of the use cases. I can go to the next one. If you have any questions, we can definitely get into it. 

Sramana Mitra: I understand what you said. As you are illustrating more use cases, highlight more about the technology in question. 

Sanjay Vyas: Let me just add a little bit of the technology portion in the previous use case as well. At the origination of the data, in any of the use cases, you need to see if you can capture that data. If you have multiple channels and multiple endpoints where the data is originating, including social media, support channels, and telephone systems; then you have multiple systems, and they don’t communicate to each other.

Now, the technology comes into the picture where the question of the problem you are solving is, “Can I bring all these data from all these different points and connect the data from customers from all these different points and get insights from it at the right amount of time when I had to make a decision.”

Before Diyotta and our technology, you would get the data from all these points, but it was not arriving in time. It takes days and weeks to bring these unmarried data and marry them. You are always doing the reactive analysis. With minimum latency in the data extraction, bringing it to the point where you can analyze it into the Big Data world and the large data ecosystem or repository, you are now able to make your decisions fast.

That is where technology plays well. It’s how you can bring it faster and bring it to the place where you can grab the insights fast before it even happens. I’ll jump to the next case in telecom, but then I’ll change the industry also. We have done a case study on this as well. There are multiple telecom fraud use cases in abuse of international calling and using data fraudulently.

When a telecom customer interacts with a device, it generates multiple records. Any phone call that you do and any data browsing that you do generates what they call a data record. It could be a call for data recall. It could be an IP data record. Multiple data records are generated.

If you multiply these data records by the number of customers, you could have millions of records in no time. The sheer volume of the data records could be petabytes within a few days. Forget about the analysis of the data. Just to get the data into a large ecosystem itself is a big issue. With the advent of Big Data and getting that vantage point, at least you have some repositories that can hold large amounts of data. 

This segment is part 2 in the series : Thought Leaders in Big Data: Sanjay Vyas, CEO of Diyotta
1 2 3