PatentAgility, KellDann’s free AI-powered patent analysis toolkit, was selected from among hundreds of applicants for presentation at an AI Builders Night at Amazon’s HQ2 in Arlington, Virginia. Kirk A. Sigmon, one of KellDann’s founding partners and the sole developer/programmer of PatentAgility, demonstrated how the platform helps patent professionals analyze claims, monitor emerging USPTO activity, evaluate examiner trends, predict likely art-unit assignment, and trace claim histories across patent families. Kirk launched PatentAgility at KellDann’s founding, and many of the tools in PatentAgility were developed as part of his graduate work at Dartmouth.
Kirk’s Presentation Focuses on PatentAgility’s Free, Comprehensive Features
Kirk’s presentation focused on the features of the PatentAgility system and its many advantages over conventional, for-profit offerings. Noting that commercially-available patent analysis tools often cost individual patent attorneys and law firms tens of thousands of dollars per year, Kirk extoled the advantages of PatentAgility: “it’s currently free, is privately hosted without use of external LLMs, we open-source it to IT teams for audits and most of it is on my GitHub, and it uses terabytes of U.S. Patent and Trademark Office data.” Presently, the system has analytics data for over seven million granted U.S. patents and over thirteen million patent application publications.
While PatentAgility has many free tools that enable patent attorneys and agents to work more efficiently, Kirk’s presentation focused on five functionalities of particular interest to the crowd:
- Claim Analysis using Natural Language Processing – Using NLP techniques and complex rulesets and standards to evaluate claims for 35 U.S.C. § 112 issues such as clarity issues, antecedent basis errors, and more;
- Semantic Alerts – Processing weekly USPTO gazette data, vectorizing that data, and running semantic search techniques on the data to alert interested parties as to cutting-edge developments in patenting activity;
- USPTO Examiner Statistics – Monitoring live patent application activity, including Office actions, patent grants, and more, to provide decades’ worth of statistical analysis regarding particular USPTO examiners;
- Art Unit Prediction – Using a custom-trained artificial intelligence model to, with 93% top-5 art unit prediction accuracy, accurately predict the likely art unit that a patent, claim, or arbitrary string of text will be assigned to; and
- Complex Claim History Analysis – A one-click tool for analyzing the entire trajectory of a patent family, including individual amendments across decades of Office actions and applicant amendments.
Kirk noted that he was currently exploring various improvements for the tool, including foreign language processing, better load balancing, and USPTO data cleaning, and more.
The toolkit received universal praise from attendees, such as investors and AI engineers, at the event. While numerous investors encouraged Kirk to “quit law” and focus on the tool, Kirk emphasized that the tool benefitted from being tested and refined through daily use at KellDann.
Try PatentAgility AI/NLP Patent Analysis Toolkit for Free
PatentAgility is currently entirely free and available at patentagility.kelldann.com or patentagility.com. The semantic search system is currently in beta and is available to anyone interested in testing the platform – those interested are encouraged to contact Kirk directly to enroll in the beta.