Cut through the legal fog with computer-generated summaries

SANTA CLARA, Calif. — Pity the poor law school graduate, staying in the office into the wee hours, until his or her eyes are blurry, reading through page after page of horribly boring — and horribly important — legal contracts.

You can imagine that, at some point, a young lawyer would start to fantasize about using a computer to summarize those contracts instead of burning the midnight oil.

That’s exactly what the founders of eBrevia did. The company, which launched today at the DEMO Fall 2012 conference here in the heart of Silicon Valley, uses machine learning to automatically summarize contracts and other legal documents.

“We’ve spent many nights as corporate attorneys scanning computer screens at 3 a.m. to extract and summarize legal provisions as part of due diligence in mergers and acquisitions,” founder and chief executive Ned Gannon told VentureBeat. “We have literally felt the ‘pain’ of the process ourselves.” Now, Gannon and his cofounders hope, they can help law firms and legal departments cut costs and save time by semi-automating the contract examination process.

People who use eBrevia’s web-based eDiligence Accelerator service specify what kinds of provisions they’re looking for. They upload legal documents and, in return, receive a summary chart that shows all the uploaded documents along with a more detailed template for each document, with special attention to the provisions that the user is looking for. Readers can easily move back and forth between the summary, the document templates, and the original documents, Gannon said.

The software uses machine learning to improve its summaries and charts as time goes on.

“It is exciting that the software gets smarter and smarter with the more examples it sees,” Gannon said.

Not everyone is so enthusiastic, however. Onstage at DEMO, American Express Ventures managing partner Harshul Sanghi said, “There’s a reason these documents are as long as they are … The devil is in the details. I’m not sure you can shorten that.”

True Ventures founder Tony COnrad concurred, saying, “I love their technology, but I don’t think legal is the place to use it.”

The company hasn’t settled on exact pricing yet but plans to give its customers — primarily law firms and corporate legal departments — a choice between per-document and monthly subscription fees.

After the launch of its eDiligence Accelerator product, the company plans to launch products for contract management, document drafting, and consumer applications. Whether it succeeds in doing that will depend on how well its summaries work. It’s hard to imagine natural language analysis that’s capable of outstripping the contextual knowledge and legal savvy of even the most exhausted lawyer. But perhaps with enough raw data and a well-defined subject area, eBrevia will help save lawyers time — even if it doesn’t eliminate the need for them.

Gannon and cofounder Adam Nguyen started eBrevia in July, 2011. The company is based in Stamford, Conn. Its presentation at DEMO is the prize for winning a national contest sponsored by Startup America and DEMO. It is funded by its founders, with the addition of the $25,000 Startup America/DEMO prize.

“Connecticut has a strong and growing entrepreneurial ecosystem, and we’re thrilled to represent the state in the Startup America/DEMO Competition,” Gannon said.

The founders are now seeking $400,000 in seed funding.

eBrevia is one of more than 75 companies chosen by VentureBeat to launch at the DEMO Fall 2012 event taking place this week in Silicon Valley. After we make our selections, the chosen companies pay a fee to present. Our coverage of them remains objective.

Photo credit: eBrevia. From Left to right: Ned Gannon, Jake Mundt, Adam Nguyen.