If you think lawyers have a cushy life, ask anyone who had to pore over long-drawn cases and make sense of the arguments. Lawyers will tell you how they have to read hundreds of pages manually, sometimes twice or thrice, to be be able to make sense of cases.
Identifying and simplifying this part of legal research is at the heart of legal-tech firm Legitquest. The Delhi-based start-up has built artificial intelligence (AI)-based technology to help lawyers with issues, reasoning, decisions, arguments and facts of all the judgments passed by the Supreme Court of India since 1950. This search feature, called iDraf, uses natural language processing (NLP) and deep learning to cull out relevant details of each case for the legal fraternity.
The idea of starting such a firm hit Legitquest's co-founder Karan Kalia while he was pursuing two degrees simultaneously - a Masters in Law from the University of Pennsylvania, and a business programme from the Wharton School a few years ago. It was while making a presentation as part of the programmes on what was happening in the technology and legal fields in India, that reality hit home. “(I realised) there is no technology and no innovation at all (in the legal field). We are actually using something that was used in the US some 15-20 years back," says Kalia.
Having worked as a Junior Associate in the chambers of late lawyer Ram Jethmalani for seven years before his Masters, Kalia took the idea of introducing technology into legal research, to his mentor.
It wasn't smooth sailing, given that no two legal cases are alike and that they actually have very little or absolutely nothing in common. Some cases, he explains, do not have a straightforward main issue or gist. "In terms of facts and arguments, it takes lawyers hours to pore over case arguments to figure out how to categorise a case. In that sense, building a keyword-based machine learning system for legal cases is not as simple as let's say a term sheet or standard rent agreement," Kalia says.
The firm secured seed funding of Rs three crore from Internet firm Infoedge India and venture capital fund WaterBridge Ventures last year and is now looking to raise some more funds. "We are in the process of actually talking about pre-series A, because in artificial intelligence, Rs three crore is not a huge amount because we have to have data analysts and data scientists who are really expensive. But what we have done with this (seed) money is that we have actually made everything very stable. The whole system is stable, so we can scale up now," says Kalia.
Legitquest has now developed the search engine with a “beautiful UI/UX experience”. Apart from iDraf, which is quite unique, it has two or three more features like infographics, which means that every time a case is opened, one can find out where it has been followed, relied upon, distinguished and so on. These things are extremely important for any lawyer to know. “Once we had the whole database of the case laws, we fed the algorithm with 20,000 cases. Then, we enhanced the accuracy of reading issues and facts from just two per cent to nearly 60-70 per cent by training the system to read better with AI and machine learning," says Kalia.
Apart from the iDraf feature, the firm also provides translation typing in filing of cases. Going forward, it plans to add more features such as analytics and chat bots on to the platform.
Legitquest's clientele consists of law universities including six national institutions that use the platform under a subscription model. Kalia says the platform also can be immensely beneficial for judiciary and law enforcement agencies in improving the conviction rate.
The firm now has Justice (Retd) M B Lokur and Justice (Retd) A K Sikri as part of its national advisory board.