Harry Potter fans will remember Professor Albus Dumbledore’s nifty memory reviewer - the Pensieve. Throughout the series, several characters used it to store their memories and rewatch them to derive insights. Somewhat like data science.
Taking this concept from fiction to reality, Gaurav Shrivastava, Co-founder of Zimmber, built a startup for a transparent data machine - called Pensieve - with Co-founder Prahlad Routh. The two were ecstatic about implementing text analytics in the legal domain and the ample growth of Machine Learning (ML) and Artificial Intelligence (AI).
The duo identified that legal tech could be a thing if Natural Language Processing (NLP) is used rightly, and started Pensieve in Mumbai in 2017. With a team of 10, Pensieve is now building an AI engine that has the capability to understand complex legal documents like judgements, laws, contracts, notices, and circulars to extract data and drive insights.
Prahlad is an IIT Madras alumnus with a background in metallurgy and materials. He also has a Master’s and a PhD in Materials Sciences from the Stony Brook University.
Gaurav is a second-time entrepreneur, having sold his previous startup to Quikr. Armed with lessons from his time with Zimmber, he wants to build the fundamentals of business as strong as possible, and not be fooled by vanity metrics.
"Startups require velocity, not speed. If the plan is not to be lucky, work very hard on the fundamentals," says Gaurav.
The legal complexity
The legal ecosystem is very text-heavy, and decision-making is driven by historical data. To put it simply, it generates insights from an existing knowledge repository. Legal research is extremely time-consuming, which often results in productivity exhaustion.
This gave Gaurav and Prahlad an opportunity to use AI to simplify data extraction, suggestion, and prediction in the legal domain, giving users ample time to deal with their clients. However, the team has to transform a customer's mindset and convince them that this technology will enhance their productivity.
Speaking about how their system works, Gaurav says,
"The engine works on judgment-data and contract-data to do clustering, semantic search, recommendation, and information extraction. The insights gathered through this AI engine are consumed by the front-end through APIs."
At present, Pensieve has two products - Mitra and Smriti.
Mitra.ai is an intuitive search engine that produces recommendations and helps a searcher prepare defensible arguments. It costs Rs 499 a month while a year’s subscription is for Rs 4,999. Using Mitra, users can get access to all cases from the Supreme Court. The web-based product has clocked over 50,000 users so far, says the co-founder.
Though Mitra.ai is a B2C product, Gaurav found a selling point with B2B customers as well because the engine can make actionable insights out of the customers’ repositories.
On the other hand, Smriti is supposed to be a one-stop-shop for contract automation, which is being developed since last year.
Pensieve has two revenue streams - subscriptions and asset management. Among its enterprise clients, Pensieve counts ICICI Lombard, Deloitte, Anoma Legal, and Infosys. It has also worked with K Law and Argus Partners. The team is making use of this opportunity to understand the legal repositories of the clients, with which it can better train Pensieve’s AI engine.
The road ahead
India is the second-largest market in the world for legal professionals with more than 600,000 lawyers with scope for increasing legal spends. According to CB Insights, legal tech startups in India have raised over $700 million in funding since 2011.
Additionally, the global legaltech AI market was valued at $3,245 million in 2018 and is expected to touch $37,858 million by 2026, according to Zion Market Research.
Though a nascent sector, India has several legal tech startups, including VakilSearch, Spotdraft, NearLaw, LawyeredIn, Legal Desk, Nyaaya, and iProbono.
The startup received an angel investment of $120,000, which was used for product development. Now, Pensieve is looking to raise a seed round of $1 million. The founders claims to have made a revenue of $30,000 and expects to clock $1 million in revenue this year.
The startup now wants to work on Indian judgments and contracts.
"The next step is to partner with major law firms in the country and corporates in insurance, IT and consulting. There is a potential to expand geographically in Malaysia and Singapore," Gaurav adds.
(Edited by Saheli Sen Gupta)