Cognition IP today announced the close of a $2.8 million funding round for its legal services AI. Trained with text from patent filings, its tools are made to augment human lawyers and help startups obtain patents and search existing intellectual property claims.
Cognition IP delivers intellectual property law services at a fixed rate, instead of adopting the hourly approach common among other legal professionals. An hourly pricing model has incentivized traditional law firms to avoid efficiency, CEO Bryant Lee told VentureBeat in a phone interview.
“Law firms just aren’t providing the level of service they could be by using technology. That’s a general problem that we’re trying to solve, trying to bring more technology into the legal world,” Lee said.
Cognition IP was founded in 2018 by Lee, a patent lawyer, and Rutgers University computer science assistant professor Andrew Tjang, who serves as company CTO.
The funding will be used to power additional AI-enabled resources and ways to search existing patent filings.
“One problem we’re working on is creating a template using AI — training the AI on existing sets of patents and what they look like to get all the background information and structure and then creating a template for that so that the lawyer can put in specific information about the invention,” Lee explained.
Cognition IP is the latest legal business to target AI that reshapes how legal professionals do their jobs.
While companies like LexisNexis have tested automated bots to assist legal professionals, startups like Trellis are making AI using the text of legal decisions to predict payout of class action lawsuits and how judges will rule in a case. Earlier this year, Disco raised $83 million for AI to assist lawyers in the discovery process.
Cognition IP has offices in San Francisco and Washington D.C. and currently has 12 employees.
The $2.8 million seed round was led by Khosla Ventures, with participation from Mayfield Fund and Basis Set Ventures. Cognition IP took part in the Winter 2018 batch at Y Combinator.