Eve, a startup that develops a personalized artificial intelligence assistant for legal professionals, announced the launch of its service today along with a $14 million seed funding round led by Lightspeed Venture Partners and Menlo Ventures.

Eve Chief Executive Jay Madheswaran told SiliconANGLE in an interview that the objective of Eve was to build the first AI hire for legal firms. He said it’s essentially an employee who could come in and immediately do assist with work alongside other lawyers with a set of prebuilt skills and learn to work within the law firm’s structure from day one and even be trained in the culture makes them stand out.

Eve began in 2020 with small language models, trying to better understand how to help legal professionals with their work, which requires a lot of day-to-day work with bulk reading, interviews and case law. Now the emergence of large language models, which power popular chatbots such as OpenAI LP’s ChatGPT capable of conversational responses and producing essays and poetry from text prompts, or producing insights from large documents, has opened up a whole new path for legal AI use.

“If you look at what lawyers are actually spending time on, it’s reading massive amounts of text. And what matters critically for them is the accuracy of the information they’re pulling out,” Madheswaran said. “How do you work alongside them when their work changes a lot every day? It’s very unstructured. This ultimately led to the birth of Eve.”

Eve differentiates itself from other services on the market, such as Harvey AI, because it’s easier to implement and focuses on personalization, Madheswaran said. As a result, smaller firms and solo practitioners can “hire” Eve and use the service as part of their casework and teach it to become part of the firm’s infrastructure.

“You can take it and teach it how you operate your firm and you can embed your secret sauce of what makes you delight your clients, what makes you process cases faster,” said Madheswaran. “Even teaching and onboarding it as you would a new hire. And that’s what we’re seeing really resonate with the market.”

Out of the box, the AI comes with an already set “skill library,” what would be expected of any new hire that includes a large number of low-level tasks covering litigation, transactional law and other specific practice areas. All of these can be easily customized by the legal firm that has taken on the implementation of Eve with their own specialization and over time Eve will learn to operate alongside the firm’s already existing employees and help them do their work faster.

Madheswaran said that the vision for Eve for most legal customers is that if they’re currently drowning in tedious “grunt work,” and they’re looking at not being able to hire a new attorney or paralegal but that’s outside their budget, they can use the AI service. It can not only become part of their operation, but it can also coach them in how to use it better.

The service is already in use by a few law firms, including Frontier Law Center, who have been putting it to use to make swifter work of their caseloads.

“With Eve, what would have been three days of work, we completed in 45 minutes, my clients couldn’t be happier,” Manny Starr, a managing partner at Frontier Law Center. “I am now able to close cases five times faster.”

As for avoiding potential problems with hallucinations and inaccuracies, Madheswaran said that Eve is built with accuracy and engineering in mind to reduce problems by grounding it. Part of that is from deep machine learning expertise and specialized training and the other part is also providing citations and accurate information to the end user. The AI also validates data so that they don’t end up with false citations, such when a lawyer from New York used ChatGPT to create legal briefs in May and the AI cited fake cases.

Madheswaran said that the funding will go to investing further in machine learning talent and growing the engineering team to continue work on growing the product and its features.

Image: Eve