Subscribe to our Feed

Thought Leaders in Big Data: Interview with Billy Bosworth, CEO of DataStax (Part 4)

Posted on Monday, Feb 11th 2013

Sramana Mitra: I would like you to do two more use cases. Feel free to pick whatever would be a good illustration of your functionality.

Bill Bosworth: What we tend to focus on is the following: When we work with larger companies, many of them have lines of business that are trying to fundamentally change the overall footprint of their business. In other words, they want to either capture new market share or they want to stop the erosion of market share from other startups. To do that, they have to move very quickly into new spaces, and they have to do that against a world that is moving very quickly around them. Often, inside of a larger corporation, our customer will be a particular line of business.

This is the case of Adobe. I have been using Adobe for many years. I am a photo hobbyist, so I think of Photoshop, Lightroom, or the Flash Player that I have used for years. Adobe also had their creative suite that handled all of their ability to build websites. But they also had to transform that into the new world. They have a new line of business that they call the Adobe Marketing Cloud. In the Adobe Marketing Cloud, they are doing really interesting stuff. The Adobe Marketing Cloud can help customers fundamentally transform how they interact with their own customers on their website. Let me give you an example: Suppose you and I own a restaurant and we have a sign-on for our loyal customers who come there. Maybe we even have a reservation system so people can come in and make reservations and order gift cards. If we would turn to a technology like Adobe Marketing Cloud, we would get to change the our customers’ experience with our website.

For example: Let’s suppose you sign on to our restaurant website as one of our customers. What we can now do is know a bit more about you. Now we understand that you went to Sonoma last week. While you were there, you visited a couple of wineries, and you loved it so much that you took the time to tweet about it. Then you went on Facebook and you liked their page and talked about it even more with another of your friends. But we also know, from your past buying patterns on our website (thanks to the analysis of the Adobe Marketing Cloud), that you tend to buy things from us between eight and ten at night on a weeknight. Now we know that you just had a great experience with a certain type of wine, we know what time of the day it is right now; it happens to be 9 o’ clock on a Thursday, and we know from all the other data about you that perhaps you would like to go out and frequent those restaurants with your friend group. What if we hit you right then and said, “Hey, would you like 10 or 20 percent discount if you come in some time over the next two weeks, and we are going to give you a complementary bottle of this type of wine?” Suddenly it feels as though we know you. It is as though we can predict what you are going to enjoy about our restaurant, and we want to make it a better experience for you.

Clients do that through using a lot of different technologies on the back end. One of them is DataStax Enterprise. They will keep anywhere between 700 million and 900 million of what they call active profiles in our system. By an active profile, I mean they need a profile where they are going to have real-time interaction with people. In that system, they service the ability to come out and tell you what things you like or what things you don’t like, so that we can change your experience with us in real time. We can change the way you are transact with us on our website. Adobe Market Cloud now becomes an enabler of taking all this data, not just from your interactions with us in the past, but from what they call second and third level data. The data about how you conduct your life in your virtual world. We can bring that together to make a more pleasant experience when you visit our website.

It is really the key to the future. The future is going to be about how to take and handle massive amounts of data of millions of people and change the experience for each of them, when they are interacting with you, to make them feel special and to make them feel that your whole company was built to service just for them. That is what Adobe Marketing Cloud does. It takes a phenomenal step in helping customers enable that experience change.

SM: So, we did two completely different use cases. The metrics use case and the Adobe use case. They are quite different.  Could you give us a third use case that is different from these?

BB: We have taken a media and entertainment use case – what I call social use cases. Let’s go to a retail use case. Ebay, for example. They have about 97 million users and about 200 million products. They are challenged with this massive amount of data in a way that can also help you easily transact on their website and make you feel like they understand you and connect you with the right people in the right places. With our systems, DataStax Enterprise, eBay comes in, take your experience around the like functionality (when you like a product), the want functionality (when you say you want a product) or your favorites (when you pull up your favorites) on a product page those are all sorted and transacted through our systems. That is a couple of hundred terabytes. They keep a couple of hundred terabytes through our systems, through that type of analysis, and serve that type of interaction. But – this is one of the neat things we bring to the table beyond our Cassandra abilities – they start to do back-end processing.

I talked about the analysis, looking for trends and patterns. What we take in that technology – in this case we take an Apache deduplication – and we have deeply integrated on top of Apache Cassandra as part of DataStax Enterprise. This way you can now have the same database in one part of your cluster, just a different part. You can now start learning that deep analysis of your data. In the case of eBay, they just built a big graph database. This means they have to show all the interconnections of the people who want stock, the people who own stock, and the products out there that people may like. They understand how all those link together. That is what they do in one big part of our cluster. They do this across data centers, and that data becomes immediately available to the real-time site, so when the user interacts with the system and needs something very quickly in response, it is going to come out of our system. If they use these different technologies within the company for those pieces of their applications, they use DataStax Enterprise.

This segment is part 4 in the series : Thought Leaders in Big Data: Interview with Billy Bosworth, CEO of DataStax
1 2 3 4 5 6 7

Hacker News
() Comments

Featured Videos