Generative AI is coming for IP and patent law, as Solve Intelligence raises its Series B

With most intellectual property or patent law work done with spreadsheets, docs, and PDFs, this is a ripe arena for generative AI, which can now address its sheer complexity better than earlier systems. There are several large competitors in the legal tech space, such as PatSnap (raised $351.6M in total); IPRally (€12.4M); and HarveyAI (which has raised over billion). 

The question is, will they survive?

Solve Intelligence is a largely UK-based startup (but now with a Delaware topco and international offices) that is applying AI to intellectual property and spitting out services for law firms. Back in April, it raised $12 million in a Series A funding round, led by 20VC.

It’s now raised $40M in Series B funding led by Visionaries fund with existing investor 20VC co-leading and “significantly increasing its stake”. Other existing investors, including Thomson Reuters and Y Combinator, also went back in, with Operator Collective and others participating. Angel investors include the founders of Tinder, Canva, Deel, Ironclad, Base44, Cleo, Hugging Face, Pigment, and Higgsfield.

The Series B brings total funding to $55M. The company claims ARR has “grown more than tenfold to reach eight figures” and that it’s profitable.

With that cash, Solve will be launching a new product for generating patent claim charts.

It now has about 400 IP teams using the platform, and customers include DLA Piper and Perkins Coie.

Over a call, CEO and co-founder Chris Parsonson told me about 60% of the customer base is big law firms, and then 40% are big enterprise technology companies and in-house teams inside companies like Siemens.

“Our software helps with drafting patent application responses, things called ‘office actions’. And the new product we're launching is for things like case litigation. So essentially, what we're solving is building specific software that helps with the patent process, which you can't use generalist legal tech tools for at the moment, so you can't use Harvey AI, for instance,” he said.

“Every patent, is defined novel. So in order to be able to go after a patent application and get it granted, and then also do stuff after patents have been granted, with licensing litigation and so on, you have to be able to understand the technology and all the new technologies, by definition. And so you have to be able to handle all these sort of complex new inventions. And you also have to be able to do it across all technology types,” he said.

He said competitors mainly do search, but Solve “does things you just couldn't do before we had this big step change in the performance of large language models… We're replacing all these workflows that were previously done inside Microsoft Word, email, and PDF previews.”

In recent times, some legal tech startups have had problem, such as Robin AI. I asked him what he thought this was down to?

“There's a huge market and a lot of people coming into the space and trying to build software for legal professionals. But of course, what happens is it's non-trivial to build the software. You have to combine legal expertise with AI researchers with software engineers to be able to build out things like evals, customize these models and algorithms specifically for legal professionals, and then also tie that into an end-to-end software package that has great UX and fits into the workflows, integrates all their existing systems, and so on. And there's no’t that many teams that can do that.”

I also asked him if he thinks VC’s are using Series B round to do “king-making” of their startups.

“It's not really something we're thinking about much,” he said. “There's basically two reasons why we've raised the Series B. We want to be aggressive on trying to scale out our technology to help make sure that as many firms and in-house teams as possible get the opportunity to at least try it. So that's why we're opening up the New York office. And also, you know, Germany's our second biggest market outside of the US, so that's why we're opening up the office in Munich.”

“The other reason is that the last two years, we've gone very deep on the drafting prosecution products, the draft and plan applications, and then the prosecution product for response, office actions. And we've got a lot of core building blocks that we now need to expand across the entire end-to-end patent process, so things like task litigation, infringement analysis, validity challenges, freedom, topic analysis and so on. So we’re doubling down on our R&D spend and basically increasing our bandwidth to be able to build across the entire end-to-end process.”

Ok so maybe he thinks VCs are doing “king-making,” but there’s clearly some evidence of this.

Legal tech reached $3.2 billion in 2025 according to Crunchbase data, but buyer demand suggests there's plenty of revenue to be had.

Stockholm-based Legora has raised $150 million in Series C funding at a $1.8 billion valuation (the round was led by Bessemer Venture Partners with participation from ICONIQ, General Catalyst, Redpoint Ventures, Benchmark, and Y Combinator).

Milan-based Lexroom has raised $19M in a Series A. Clio, based in Vancouver, raised $500 million at a $5 billion valuation and acquired vLex for $1 billion. Zurich’s DeepJudge raised $41.2 million at a $300 million valuation.

Other recent fundraises include SpellBook ($50 million), EvenUp ($150 million at $2 billion valuation), and Eve ($130 million at $1 billion valuation).

Company data: