Stilta Secures $10.5M Seed Round to Automate the Drudgery of Patent Law
Table of Contents
Turning Dormant IP into Active Assets
For most large corporations, the patent portfolio is a graveyard of expensive ideas. Millions of dollars are spent filing intellectual property (IP) that eventually sits on a digital shelf, unanalyzed and unenforced, simply because the cost of auditing those patents—and the litigation required to defend them—is prohibitively high.
Stilta is betting that AI agents can collapse that cost barrier. The startup announced a $10.5 million seed round on Tuesday, led by Andreessen Horowitz (a16z), with participation from Y Combinator and strategic investors from OpenAI, Legora, and Lovable. The goal is to move patent research from a manual, weeks-long slog into a high-speed analytical process.
The company was born from a realization that the legal profession is still operating on analog-era workflows. CEO Oskar Block, who previously focused on machine learning for sports betting and AI integration consulting, found inspiration in a conversation with co-founder Tobias Estreen. Estreen’s father, a veteran patent attorney, described a career spent reading essentially the same types of documents for three decades. That redundancy, Block realized, was a massive market opportunity.
How the Agentic Workflow Functions
Unlike simple LLM wrappers that summarize a document, Stilta employs a network of AI agents designed to mimic a room full of legal specialists. When a user inputs a patent number and relevant context, the system doesn’t just search for keywords; it reasons across datasets in parallel.
The agents are tasked with identifying conflicting patents, flagging similar intellectual property that could jeopardize a claim, and excavating deep filing and court histories. The result is what Block describes as “litigation-grade” output: comprehensive reports and claim charts featuring pinpoint citations to evidence, allowing a human lawyer to remain in the driver’s seat without doing the manual heavy lifting.
This approach targets the “analytical bottleneck” that defines modern IP law. By automating the discovery phase, Stilta allows companies to rediscover the value in patents they forgot they had, potentially turning a dormant portfolio into a source of licensing revenue or a defensive wall against competitors.
The Shift in Legal Tech
Stilta enters a competitive landscape alongside players like DeepIP and Solve Intelligence, arriving at a moment when legal tech is seeing a surge in venture interest. While the broader legal industry is often slow to adopt new technology, the data-heavy side of the house—specifically IP and discovery—is proving to be the easiest point of entry for AI.
Block notes that while the final decision-making process in a courtroom will remain human-centric for the foreseeable future, the work leading up to that moment is already being overtaken by automation. The shift isn’t just about speed; it’s about accessibility. When the cost of analyzing a patent drops, companies are more likely to enforce their rights and more cautious about infringing on others.
The funding comes at a time when enterprises are moving past the “experimentation” phase with AI and are looking for specialized tools that solve high-value, narrow problems. For Stilta, that problem is the millions of pages of patent filings that currently require a human eye to make sense of.
