From left, Andrew Arruda, Akash Venkat, Pargles Dall’Oglio, Jimoh Ovbiagele and Shuai Wang developed an artificially intelligent legal researcher named ROSS.Kevin Van Paassen/The Globe and Mail
A class project-turned-startup launched by University of Toronto students that uses IBM's artificially intelligent Watson computer to do legal research now has backing from Dentons, the world's largest law firm.
Called Ross, the app uses Watson, which won the TV quiz show Jeopardy! in 2011, to scour millions of pages of case law and other legal documents in seconds and answer legal questions. Its founders liken it to a smarter version of iPhone's Siri, but for lawyers, and say it could one day replace some of the grunt research work now done by low-level associates at the world's top law firms. It is one of several attempts to apply what is called "cognitive computing" to the historically technology-averse legal profession.
Dentons, the recently created global law firm that encompasses Canada's former Fraser Milner Casgrain LLP, has announced an undisclosed investment in startup Ross Intelligence Inc. The company will make use of the law firm's NextLaw Labs, a project aimed at developing new technology for the legal business and based in Palo Alto, Calif., in the heart of Silicon Valley. It is the first publicly disclosed startup that the NextLaw Labs project has signed up.
Dentons also announced a partnership last week with International Business Machines Corp., based in Armonk, N.Y., to provide legal startups working within NextLaw Labs access to a technology platform using IBM's cloud computing resources.
Ross, still in the testing phase, is being piloted not just at Dentons but in a list of elite U.S.-based law firms, its founders say. The program, using a knowledge base focused largely on U.S. bankruptcy law, is being used but is still being refined.
However, the boost from being connected to the world's largest law firm is undeniable, co-founder Andrew Arruda said, as the startup develops the tool into something law firms will be able to use off the shelf.
"It's early days for sure," Mr. Arruda said. "But what we are seeing is Ross grasping and understanding legal concepts and learning based on the questions and also getting user feedback. … Just like a human, it's getting its experience in a law firm and being able to learn and get better."
Joe Andrew, the global chairman of Dentons, said in an interview that the investment in Ross was exactly the kind of project NextLaw Labs was meant to take on, a promising technology that needs a law firm onside to help develop it into something lawyers will actually use: "We all recognize that the biggest problem for technology and the law is to get lawyers to actually adopt it."
Other similar legal tech incubators have cropped up in recent months. Two have recently opened in Toronto: LegalX, a "cluster" located at the Ontario government's MaRS Discovery District, and the Legal Innovation Zone (LIZ) at Ryerson University.
The MaRS project is home to a legal startup called Beagle, which also has support from the Seattle-based Microsoft Ventures Accelerator program, which uses artificial intelligence to sift through and explain the important parts of complex contracts. Ryerson's LIZ houses two startups creating legal websites that act as "dating services" for lawyers and clients.
Ross also has backing from Y Combinator, a U.S. seed fund that has backed several tech startups that have become household names, including accommodation website Airbnb and social-media website Reddit. As with Dentons, Mr. Arruda declined to say how much money Y Combinator has invested.
Ross, which is less than a year old, was a project launched at the University of Toronto last year to compete in a contest that saw IBM offer its Watson platform to teams at 10 universities – all but U of T in the United States – for the development of commercial applications. Ross faced off against the nine other schools' teams in New York in January. The U of T students missed out on the $100,000 first prize, but kept their access to Watson and support from IBM and have been developing their concept ever since.