IPRally was founded in 2018 in Helsinki, Finland, with an approach to patent search based on graph AI rather than keyword or semantic text matching alone. The company raised €2 million in seed funding and followed with a €10 million Series A in March 2023 led by Endeit Capital. IPRally has been recognized as one of Finland's fastest-growing technology companies by Deloitte's Technology Fast 50 and has built its platform on Google Kubernetes Engine with Ray for distributed processing at scale.
IPRally's search engine converts patent documents into graph representations that capture the technical relationships between concepts described in claims and specifications. This graph-based approach allows searches to be based on technical meaning rather than keyword overlap, and results are returned with transparent, adjustable graph queries that show why each result was surfaced. The platform also supports image search — users can upload photographs, technical drawings, or flowcharts and the system decodes their technical content to find relevant patents. Coverage spans 58 jurisdictions and over 120 million patents, including machine translations for non-English documents.
IPRally serves patent searchers, IP attorneys, and R&D teams who conduct freedom-to-operate, novelty, state-of-the-art, and invalidation searches. The platform is positioned for professional search use rather than casual browsing, with tools for building comprehensive search reports and managing classification workflows. Explainability features — showing the graph structure behind each result — are designed to support the audit trails that IP practitioners need when documenting search methodology.
IPRally's graph AI approach is its primary technical differentiator. Traditional keyword and semantic embedding approaches return results based on linguistic similarity, whereas IPRally's graph structure captures functional and structural relationships between claimed inventions, enabling searches that surface technically similar patents even when written in very different language.
Hands-on review pending.