EVRI’s strategy to solve the information overload problem for consumers at the point where the consumer wants to engage is beginning to bear fruit. Following on the heels of its partnerships with established content provider organizations such as The Washington Post, it has launched in partnership with the Hearst Entertainment-backed site LMK.com semantically-enriched portals to the worlds of college football and most recently reality TV.

CEO Will Hunsinger of EVRI says expect to see the strategy continue to play out in more deep partnerships like this one. “We don’t want to be a powered by company,” he says. “We want to reach consumers with a more compelling way to find out what the web is talking about, what is important in the real-time web, not just be a technology provider.”

Portals such as LMK.com will draw from the superset of technologies that EVRI develops for its own destination site, which Hunsinger says is evolving into EVRI V2. It has re-categorized the home page with multiple popular streams based on different consumer categories, from sports to politics, and it plans to continue innovating on its core mission: Using semantic technologies and natural language processing to understand and slice and dice that real-time stream of data currently being shared and discussed, and packaging that for ease of consumer consumption. More emphasis is being placed on entities within that content and how they are interact with one another within the context of something that’s happening – Tiger Woods, Florida car crash, and Elin Nordegren Woods might have some interest as topics on their own, but where things get interesting is of course where those worlds collide, and those are the intersections where EVRI wants to play.

“We also feel like our core is pretty strong so we’re going to start to introduce some more social inputs into relevancy,” Hunsinger says. “We know people are talking about something and our technology sees the intersection of entities across sources. But if you introduce the velocity aspect, how quickly it’s being shared and consumed-- maybe even a popularity feed based on engagement--that’s when we feel we can really start to tighten the relevancy of all of it in real time.”

Hunsinger comments that the start-up continues to work deeply at applying different semantic lenses on top of information – including the sentiment lens, which is a beta project. That works well now for the people, places and things heavily discussed on the web, he says, and he expects EVRI to get better at it. Also in his sights is identifying and delivering content on a more conceptual or event-based basis. “I’m using the term ‘event’ not like someone’s birthday, but in the sense that we see Ben Bernanke, Timothy Geitner, and TARP as entities, and Fannie Mae and Freddie Mac, and then building velocity and volume of content and extracting and delivering a conceptual event, like the housing crisis, as an emerging means on the web.”

That’s a hard nut to crack, he says, and EVRI is taking it step by step, with the hopes of having something see the light of day in 2010. “It’s a preview into where we are going in terms of using semantic technologies in a different way,” he says. It’s experimenting now using a corpus of content it acquired that was hand-categorized into different conceptual buckets. “We’re training algorithms to see if we can understand the pattern recognition of how they were editorially curated intp conceptual buckets,” he says, “to train our system to identify that same type of categorization in an algorithmic way.”

Over the next few months he also expects EVRI to expand its presence on mobile platforms, and to deliver a desktop alert app on the people users are following across the web, in the style of TweetDecks but based on semantic technology rather than the social graph.