Our Take
A16z is betting on agentic workflow automation in litigation, but Stilta's $10.5M entry into a category where Patlytics ($65M raised) and Solve Intelligence ($40M Series B) already operate tells you product fit, not market newness, is what attracted capital.
Why it matters
Patent litigation teams spend months on invalidity searches and claim mapping. If Stilta's agents can materially cut that cycle, in-house IP teams and mid-market law firms will adopt it. The category has compressed from zero to three serious contenders in six months.
Do this week
IP counsel: request a Stilta demo this week against a patent you're currently analyzing so you can compare recall and precision against your current search workflow before budget season.
Stilta Closes $10.5M Seed Led by Andreessen Horowitz
Stockholm-based Stilta, founded in December 2025 by four former McKinsey engineers, announced a $10.5 million seed round led by Andreessen Horowitz general partner David Haber. Y Combinator, which backed the company earlier in 2026, also participated, alongside founders from Sana, Legora, OpenAI, Lovable, and Listen Labs. This is Stilta's first outside funding since launch five months ago.
The company formally shipped its product in February 2026. According to the press release, it has already signed enterprise customers including Roche, Alfa Laval, and Maersk, along with three of the world's five largest IP litigation firms as either customers or active pilots (specific names withheld by Stilta).
Litigation Work, Not Prosecution, Is the Wedge
Stilta enters a crowded field. Patlytics closed a $40 million Series B led by SignalFire in April 2026, bringing total capital to roughly $65 million. Solve Intelligence, based in San Francisco and London, raised its own $40 million Series B in December 2025 and launched Charts, a product targeting the same invalidity and infringement workflows Stilta pursues.
The difference: Stilta's founders are positioning litigation as their moat. Most existing patent-AI tools focus on prosecution (drafting claims, responding to office actions) rather than litigation (hunting prior art, mapping claim scope). Patent litigation workflows are labor-intensive and have not substantially changed in decades, Haber noted in the press release. Stilta automates them by dispatching AI agents to search across 180 million patents, 250 million scientific publications, and more than a trillion archived web pages via integration with the Internet Archive's Wayback Machine.
In a demo, Stilta's CEO Oskar Block showed the system surfacing 868 prior art references in roughly half an hour against a wireless-networking patent, color-coded by strength of read against each claim limitation. The platform also supports infringement and freedom-to-operate analysis on the commercialization side. Block claimed internal benchmarking shows Stilta achieves roughly three times the recall of general-purpose models like ChatGPT, Claude, and Perplexity on invalidity tasks (no independent verification disclosed).
The customer split is two-thirds in-house IP teams (pharma, industrials, high-tech) and one-third law firms. Roughly 50 percent of the base sits in the United States; 50 percent in Europe.
What Practitioners Should Watch
Stilta's long-term pitch mirrors Patlytics and Solve: become the "full IP operating system" spanning litigation, prosecution, and adjacent workflows. Block argues that starting in litigation, where accuracy demands are highest, makes expansion to lower-stakes work easier rather than harder.
The funding will hire beyond the four co-founders, with engineering and go-to-market positions in Stockholm and a planned New York office by year-end. Stilta has not disclosed revenue, pricing, customer counts, or independent benchmarks validating the 3x recall claim.
For in-house IP teams already paying for Westlaw or LexisNexis workflows, or law firms managing multiple patent dockets, the question is whether AI-driven recall at scale materially compresses search and analysis time. The three-way competition among Stilta, Patlytics, and Solve will clarify pricing and feature gaps over the next 12 months.