- Worldwide demand for patents has skyrocketed, but the patent filing registration process hasn’t changed since the 1990s.
- NYC and Paris-based DeepIP uses trusted AI to augment IP attorneys, helping them deliver better quality patents and save up to 50% of the time spent drafting applications.
- $15M Series A led by AI fund Resonance, with participation from Headline (Mistral AI), Serena Capital (Dataiku), and Balderton Capital.
New York, Paris March 27 - In a sense, patents are the underlying foundation of capitalism. They play a vital role in driving progress, protecting new ideas, and ensuring fair competition. Breakthroughs that begin in research labs require patents to transition into commercial applications. For example, mRNA vaccines were patented by a university and licensed to BioNTech and Moderna to become COVID-19 vaccines. Without patents, most innovations wouldn’t move beyond the lab.
Unfortunately, the patent filing process hasn’t changed in 30 years. Patent applications are still drafted the same way they were—word by word, document by document. While legal tech has modernized contracts and compliance, patent registration remains trapped in a maze of manual work. DeepIP changes this with AI.
Patents Are Stuck in the Days of Fax Machines
Once an R&D team has completed research and the outcome is found valuable, it is handed over to attorneys who specialize in Intellectual Property (IP) rights. The process is very expensive, costing anywhere from $20K to $100K in lifetime costs, and slow—it might take 2 to 4 years for the patent to be granted, with over 1 million applications awaiting review by the United States Patent and Trademark Office. Additionally, the process is prone to human error. The literature is complex, the terminology is constantly evolving, and the process is low-tech, relying almost entirely on human-limited cognitive capabilities.
A mistake results in a patent that is non-defensible and therefore useless. The pressure on attorneys is immense, but IP talent is scarce (in the US, becoming a qualified patent practitioner requires 7-10 years and $200,000 in tuition costs) and they are expected to deliver applications faster than ever—down from 40 hours in the 1990s to just 20 hours today. However, the way they work has not changed. Aggravating things, as research becomes more advanced, demands continue to grow at a +4.4% compound annual rate, with 3.5 million patent applications filed worldwide.
Co-founders François-Xavier Leduc (CEO) and Edouard d'Archimbaud (CTO) have spent the past five years creating AI solutions for research-driven institutions like Airbus, IBM, and SAP. They have mastered transforming data into reliable AI—meeting and even exceeding the stringent security standards of the U.S. Department of Defense.
Trustworthy AI for Patent Practitioners
Bringing together top AI and engineering talent from leading universities with experienced patent practitioners in NYC, DC, and Paris, DeepIP serves patent professionals by aligning with the way they work. The product combines third-party models, proprietary models, and unique data access for better performance and far fewer AI hallucinations. It supports patent practitioners in their daily tasks across patent preparation, drafting and prosecution.
“There have never been as many innovations that could be protected and create more value for the entire economy. Yet, the way we manage the patents that make this possible hasn’t evolved fast enough. We built DeepIP through continuous feedback loops with the 50 largest U.S. IP law firms, refining our AI to deliver with precision and legal rigor. By combining cutting-edge AI with deep legal expertise, we’re pioneering a new era where patent professionals collaborate seamlessly with AI to work faster, smarter, and more strategically. We augment, not replace,” says CEO François-Xavier Leduc. "The product-market fit was almost immediate," he continues. "As one customer put it: 'I already have five colleagues who told me, If you don’t buy this for the firm, I will go buy it myself.'"
Harvey for Intellectual Property Practitioners
Law firms have been leveraging technology for a relatively long time, with computer-assisted legal research and document review products in use since the early 2010s. More recently, “legal-grade AI” startups like Harvey, last valued at $3B, have emerged. DeepIP is now unlocking the same value for the IP industry.
Founded in 2024 in NYC & Paris, DeepIP augments rather than replaces existing patent practitioners. The company has already helped draft 8,500 patent applications, with lawyers reporting up to a 50% reduction in time spent drafting, freeing roughly two hours of work per day. This has led not just to better productivity and improved IP firms’ talent retention, but better quality, with patents that are more robust and more defensible.
A feature that sets the product apart is its ability to analyze past patent applications, allowing DeepIP’s AI to learn and replicate the writing patterns of individual practitioners and their clients. This is crucial, as clients often require a specific tone and particular wording.
Unlike standalone AI tools that disrupt existing workflows, DeepIP is also the only AI patent assistant fully embedded in Microsoft Word, offering a seamless integration within attorneys' workflow. DeepIP enforces a strict Zero Data Retention policy and is fully GDPR-compliant, as well as SOC 2 Type II and ISO 27001 certified, ensuring complete data segregation and confidentiality.
Funding from Resonance, Headline, Serena Capital, and Balderton
DeepIP estimates the global IP law services industry to be worth $42B per year, with an immediately addressable market worth $1.3B. Today’s $15M Series A funding round is led by AI fund Resonance, with participation from Headline (Mistral AI) and Serena Capital (Dataiku), with Balderton also on the cap table. The company will use the funding to expand its operations in the U.S. and Europe, evolving its current productivity assistant toward a more agentic approach and continuing to add intelligence features.
"DeepIP is the fastest startup I’ve ever seen reach 7 figures in ARR in 7 months! To me, this is a clear testament to the massive productivity gains DeepIP is bringing to patent practitioners. I’m convinced that DeepIP will become an essential part of every IP law firm’s operations in the years to come. Having partnered with François-Xavier and Edouard on their first company, Kili, I’m thrilled to join them again on this new journey with us” says Maxime Le Dantec from Resonance.
"AI isn't replacing entire professions; it’s integrating into specific tasks, making knowledge-based work faster and more efficient. DeepIP is the perfect example of this trend: it augments IP practitioners to the point that once they try it, they don’t want to work without it. It's a B2B product, but IP lawyers derive so much value from it that they’ve told us they would personally buy it if their company didn’t. Patent filing is essential for scientific advancement and DeepIP is building its missing piece," says Jonathan Userovici, General Partner at Headline.
{{block-cta}}
.png)