Nine months. That's how long DeepIP waited between raising a $15 million Series A and closing a $25 million Series B on March 3, 2026. The compressed timeline speaks to something venture capitalists find increasingly rare: a legal tech product that lawyers not only tolerate but actually adopt.
The NYC and Paris-based startup, which embeds AI-powered tools directly into the Microsoft Word environments where patent attorneys already spend their days, attracted Korelya Capital and Serena to co-lead its latest round. Existing investors Headline and Balderton Capital joined in, bringing DeepIP's total capital raised to $40 million—a sum that reflects both enthusiasm and, perhaps, the high cost of convincing a notoriously change-resistant profession to trust artificial intelligence with billable work.
More than 400 law firms and corporate intellectual property teams now use the platform across 25 jurisdictions, including Philips, Greenberg Traurig, Dexcom, and Mewburn Ellis. The company says it has assisted on over 40,000 patent matters since launch. Revenue has grown tenfold in 18 months, according to investor disclosures, though the company declined to share absolute figures.
The Microsoft Word Gambit
DeepIP's core insight was disarmingly simple: don't make patent attorneys learn new software. The platform runs as a plugin within Microsoft Word, handling everything from invention capture and prior art searches to drafting, prosecution, portfolio analysis, and risk assessment—all without forcing users to toggle between applications or master unfamiliar interfaces.
According to March 2026 reporting from French tech publication Journal du Net, the platform cuts patent application drafting time from roughly 25 hours to around 12. If accurate, that's the kind of efficiency gain that law firm partners—who bill by the hour but compete on turnaround time—find difficult to ignore.
Still, adoption in legal hasn't come easy. DeepIP maintains SOC 2 Type II and ISO 27001 certifications and operates under a zero data retention policy, measures designed to address the confidentiality concerns that have historically made law firms skittish about cloud-based tools. Whether those credentials fully allay fears about training AI models on sensitive client work remains a tension the company navigates client by client.
Betting on a New System of Record
Korelya Capital, the Paris-based firm founded by former French Digital Minister Fleur Pellerin, sees DeepIP becoming what it calls a "system of record" for patent work—not just a tool for one task, but the default platform where patent professionals manage their entire workflow. Serena, an early backer of French unicorn Dataiku that closed a €200 million fund targeting AI and climate tech in October 2025, shares that thesis.
It's an ambitious position. Roughly 80% of DeepIP's clients sit in the United States, despite the company's dual-headquarters structure. That geographic concentration reflects both where patent work happens at scale and where firms have budget flexibility to experiment with new platforms.
Co-founders François-Xavier Leduc (CEO) and Edouard d'Archimbaud (CTO) come to the problem with relevant scars. The pair previously built Kili Technology, an AI data labeling platform. DeepIP actually began life as a patent drafting tool within Kili before spinning out and rebranding in August 2024—a pivot that suggests the founders saw more potential in legal workflows than in the increasingly commoditized data annotation space.
Expansion Mode, With Questions Embedded
The Series B will fund geographic expansion in the U.S. and Europe, though DeepIP hasn't named specific cities or market segments it plans to target. The company intends to roughly double headcount from approximately 45 employees today to somewhere between 90 and 100 by year-end 2026, with a focus on engineering, product development, and go-to-market teams.
The broader context is favorable, at least on paper. Global patent filings hit a record 3.7 million in 2024, up nearly 5% year-over-year, according to data from the World Intellectual Property Organization published in 2026. But whether AI platforms like DeepIP ultimately expand that market—by making patent protection more accessible to smaller companies and individual inventors—or simply help existing law firms service their current clients more efficiently remains unclear.
For now, DeepIP appears content to pursue the latter. Selling to established firms and corporate IP departments offers a clear path to revenue. Making patent law more democratic? That's a longer, messier bet—and one that doesn't necessarily align with the interests of the law firms currently signing contracts.
Nine months between funding rounds suggests DeepIP won't have to wait long to find out which path proves more lucrative.
