Seth Redmore is the vice president of product management at Lexalytics, a company that provides an engine to convert unstructured text into structured data. In this interview Seth talks about how Lexalytics provides big data solutions for its customers, the direction of development Lexalytics is taking, and how the company is willing to work with startups to achieve its goals.
Sramana Mitra: Seth, let’s start with some context. Tell us about Lexalytics, what you do, who your customers are, and what value you provide so the audience knows whom we are talking to.
Seth Redmore: I am the vice president of product management at Lexalytics. I have been in tech for 20 years. I started in networking and I did a couple of networking companies, including co-founding companies that we sold to Cisco back in 2000. While at Cisco, I got very interested in the problems of how to have a credible and influential marketing department. Marketing is a difficult organization to measure because there are a lot of compounding factors. If you look at sales and bottom line stuff, there are a lot of things that are affected: competition, how good the product is, the sales team, etc. If you take a step back and say, “First thing that proves our job is to change the discussion.” Then we ask ourselves, “How do we measure that?” I was very interested in PR analysis and media measurement while I was at Cisco. So I built Cisco’s reputation management system along with a team of other people. They still use a second version of that now. I learned a lot about the mechanics of doing something like this. This is why I started getting interested in the area of text analytics.
Text analysis is what Lexalytics does. In essence, it is turning unstructured text into structured data that you can use to perform calculations on and do all the stuff that computers are really good at. In other words, if you give us a piece of text, we will tell you who is being discussed in there, what the context of the conversation is, if the conversation is positive or negative, where the places are that are being discussed, etc. You take all the information about where and when and use that to figure out why and what you are going to do about it. In a nutshell, that is what we do.
SM: When you talk about text analysis, are you doing this in any specific domain, or is it just broad text analytics technology that you are applying to different problems in your licensing technology?
SR: A combination of both. The nature of text is such that you have to be a little domain conscious. There are some domains where you have to be much more conscious than others. We tend to focus most on what I would call relatively informal, generally non-technical communication. What that means is we deal a lot with media, social media, survey responses, feedback coming in, etc. What we tend to deal with less are highly technical papers on particular chemical reactions. That tends to require fairly specific technology. We have some customers who are doing that stuff, but our focus tends to be much broader and looking more at the media and survey side. If you look at social media monitoring companies, we power out of the ones that haven’t built their own engine. We power most of the top four or five monitoring companies.
It’s the same in survey analysis. In this space we are the number one OEM engine provider. Then we have presences in companies like pharmaceuticals to do diverse event tracking. A doctor calls in and says, “This patient had this particular side effect.” So we do tracking and roll-ups for manufacturing companies. A couple of industries that are starting to come online now are cyber intelligence and e-discovery.
SM: If you look at your customer base today and your business, what is the bulk of the business? Is it an OEM engine business?
SR: Yes. I consider us kind of a weapons manufacturer. We build an engine and we sell it to a bunch of people who then customize it as they wish. But we specifically did not get into the business of gathering content or getting too much into the analysis part because we do not want to compete with the people using our engine.
This segment is part 1 in the series : Thought Leaders in Big Data: Interview with Seth Redmore, VP of Product Management at Lexalytics
1 2 3 4 5