Adaptive Semantics is an online customer management system which can be customized for different publishers. This company based on Linguistic Algorythm was founded in November 2008 by Elena Haliczer and Jeff Revesz. The NextWomen interviews Elena Haliczer on creating innovation and founding the business, and its recent acquisition by The Huffington Post.

"The Huffington Post made our decision very easy. Not only was it an attractive offer that felt right at the time, it was also a tremendous opportunity to work for a company with a real track record for innovation."

How did you come up with the idea of your start-up? And what is its USP?

In the early days of our company, Jeff Revesz, my co-founder, and I felt strongly that the market really made a lot of our decisions for us. When we  first started talking about forming a company, there was a lot of chatter and frustration about the lack of any really effective comment moderation system for blogs. At the same time, there was a lot of interest around discovering the influencers within online communities. Jeff had already begun working on JuLiA (Just a Linguistic Algorithm) in his spare time, and we were looking out for a good application for the technology. We  quickly realized we could use JuLiA as the basis for an automated comment moderation and expert discovery system and market it to publishers.

We use linguistic algorithms and text data analysis -- which become more sophisticated over time -- to help moderate comments by recognizing offensive language and patterns of behavior used to evade comment moderation.

While publishers often rely on filters to moderate comments, they don’t effectively catch all the content necessary as users learn to work around keyword filters. Given the volume of comments that many publishers receive and the growing importance of comments to publishers' communities, relying on human moderators can become unmanageable. The linguistics algorithms of our technology get around these issues by sorting through extensive databases -- a fast-growing list of hundreds of thousands of single words and keyword combinations -- to eliminate abusive comments. Through data mining and statistical analysis, JuLiA improves the user experience by using analytics to recommend community experts and identify and eliminate abusive posts.

We focused our product on publishers from the very beginning, and at the time, there simply wasn't anything like it out there. Sentiment analysis itself had not really entered into the public conversation. This changed over the course of 2009 as companies like Crimson Hexagon and Jodange came to the forefront. We had some especially good timing, because we were already out there with clients who had gotten good results with JuLiA. Sentiment analysis suddenly became this class of products that publishers understood and wanted.

Companies that were potentially our competitors actually made our sales a lot easier.

How did you form your team?

Jeff and I met at New York Internet Week in June, 2008. He was working at AIG, and I was working at Games for Change, a nonprofit focused on social issue games. We were both fishing for a potential co-founder, though at the time neither of us were completely sure what the product would be. The match felt right because he had a strong background in math and I had an equally strong background in marketing, but more than that, we were both developers, and both hungry for the startup life. Concrete ideas and a minimum viable product were developed over the course of 2008, and in November we officially founded Adaptive Semantics.

How did you fund it, with how much money, and what is the business model?

Without specifics, we started Adaptive Semantics with a combination of our own savings and a small friends and family investment.

In April, 2009, The Huffington Post made a strategic investment in Adaptive Semantics. JuLiA was a B2B product that we sold under a tiered subscription model. At the time of our acquisition we were profitable and working with a number of tier one publishers. We had also begun previewing a series of new products based upon the core technology.

Why a sale (already)?

When you start a tech company, your motivations are many, but what you dream of is having the opportunity to really innovate, to completely rock the tech.

The Huffington Post made our decision very easy. Not only was it an attractive offer that felt right at the time, it was also a tremendous opportunity to work for a company with a real track record for innovation.

What has changed since the sale for you and your business?

The biggest change has simply been having a team of developers to work with. It is also a lot easier to focus on what we enjoy most, which is product development.

How did you approach and manage the sale? Did you talk with more parties etc.

We approached the sale the way we approached every major business decision we made.

We consulted with each other and our advisors, took the time to meet with lawyers and discuss the legalities, and did what the felt best for both of us.

What was your biggest challenge during the development process and how can other start-ups learn from that?

Deciding where to put your focus is the biggest challenge you face in the development phase. Most of your time is spent on development, but you also have to get out there and pre-sell, network, and find funding. I learned that it is okay and actually quite important to get out there and introduce yourself to potential investors even if you don't have a minimum viable product or sales to show off. Talk about your idea, email them, post comments on their blog, pre-sell yourself and your product. And if you get rejected anyway, keep in contact. Send notice of your successes, introduce them to other entrepreneurs that might fit their investment model better than you did.

Are you profitable?

We were profitable at the time of the acquisition.

Where is the company in 1-5 years?

We were acquired, and are currently working at The Huffington Post.

After our tenure at HuffPost, I plan on starting another company.

What is your view on pay walls/ free content?

The freemium model works for apps. It remains to be seen if it will work for publishers. I do think personalized news, custom ebooks, news packaged with social content from Twitter and Facebook are all content models people are interested in and will probably pay for. However, simply putting up a paywall strikes me as an action guaranteed to alienate a significant segment of your readers.