Stilta raises seed to automate patent analysis with AI agents
Stilta raised a $10.5 million seed round led by Andreessen Horowitz to build agentic AI software for patent enforcement, defense, and commercialization. Business Insider reported July 8, 2026, that the startup was founded in January 2026 by former McKinsey consultants and uses agents to search patents, academic papers, file histories, web archives, and other evidence sources. For practitioners, the important technical problem is not simply summarizing patents; it is high-recall retrieval, claim-level provenance, and auditable evidence mapping in a legal workflow where missed sources can change case strategy.
Industry context
Patent analysis is a useful enterprise-AI test because the task is retrieval-heavy, source-sensitive, and expensive when done manually. Stilta's pitch is not just that an agent can read legal material; it is that the system can connect evidence to specific patent claims with enough provenance for lawyers and corporate teams to trust the result.
What happened
Business Insider reported on July 8, 2026, that Stilta raised a $10.5 million seed round led by Andreessen Horowitz. Stilta's own site says the round was announced on May 19, 2026 and describes the company as building agentic AI for patents. DLA Piper and Legal IT coverage also described the investment and said the product targets patent enforcement, defense, and commercialization.
Technical context
The hard engineering work is high-recall retrieval across patents, papers, file histories, web archives, and internal evidence, then mapping findings to claim-level arguments. For ML teams, that means provenance metadata, auditable search logs, and human review are product requirements, not optional polish.
What to watch
The next proof point is whether Stilta can show repeatable legal outcomes, not only faster research. In this category, a model that misses a key reference or cites weak evidence can create costly strategic errors, so evaluation should focus on recall, citation quality, and lawyer acceptance.
Key Points
- 1Stilta raised $10.5 million to build agentic AI workflows for patent enforcement and defense.
- 2The technical challenge is high-recall retrieval plus claim-level provenance, not generic legal summarization.
- 3Practitioners should prioritize auditable search logs and human review when applying agents to legal evidence workflows.
Scoring Rationale
The funding is notable because it applies agentic AI to a high-value legal workflow with strong provenance requirements. It is not major platform news, so the score remains in the solid-to-notable range.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
