A few weeks ago, I came across a Seattle area startup called Wavii that had listed a job opening on nPost. I'd never heard of Wavii, so I wanted to find out more. I posted an inquiry on Twitter and reached out to the company's founder, Adrian Aoun, who wasn't quite ready to talk.

"We’re currently trying to stay quiet about what we’re up to, and thus would appreciate as little press as possible…but we will be talking more about it in the coming months," Aoun said in the June 4th email. Further inquiries turned up very little about what Wavii was actually building.

But now TechCrunch has dug up a little more info about the stealthy Seattle startup, though the entire project still appears to be shrouded in mystery.

TechCrunch -- sourcing its report on "spurned venture capitalists who've desperately wanted to invest" -- says that the company is working on Natural Language Processing technologies that are designed to tie bits of information together in new ways. The service eventually will automatically create "social content in real-time from news, blogs, and tweets across the web," reports TechCrunch.

That sounds a little bit like what the highly-touted Flipboard -- which just emerged from stealth mode this week -- is up to. And it also ties into what Seattle-based Evri has been building over the past few years.

In fact, one of the listed employees at Wavvi is former Evri developer Michele Banko, who recently received a PhD from the University of Washington working with noted computer science professor Oren Etzioni. Etzioni, who previously founded Farecast and Netbot, is a venture partner at Madrona Venture Group. (Here's a paper that Etzioni and Banko co-authored in 2007 titled "Structured Querying of Web Text.")

And, perhaps more interestingly, here's a look at Banko's PhD dissertation from 2009, which certainly parallels with what TechCrunch is reporting. The first paragraph of Banko's dissertation -- titled "Open Information Extraction for the Web" -- notes:

"The World Wide Web contains a significant amount of information expressed using natural language. While unstructured text is often difficult for machines to understand, the field of Information Extraction (IE) offers a way to map textual content into a structured knowledge base. The ability to amass vast quantities of information from from Web pages has the potential to increase the power with which a modern search engine can answer complex queries."

It is unclear whether Banko's work at the UW is forming the backbone of Wavii, but it wouldn't be the first time that a UW computer science project found its way into a startup company. Meanwhile, TechCrunch reports that suitors have already lined up to buy Wavii, including Yahoo which reportedly offered up to $10 million in cash.

We'll be tracking Wavii as more details emerge, and if you have more information about what they're building or who is involved please feel free to share.