The IPRally engine contains three basic elements. First, a graph-based data model which represents how patent attorneys and examiners perceive things when drafting or examining patent applications. Second, the prosecution history of earlier real patent applications. Third, a cutting edge neural network that learns the patenting logic using these. An intelligent search engine that thinks like a patent professional is born.
Read more from the blog
IPRally combines user control, speed of searching and unmatched search accuracy:
• Visual and intuitive user interface
• Proven relevant hit finding performance
• Search time about 3 seconds
• Built-in semantic and technical understanding
• Automatic relevant passage highlighting
• Iterative search
• Competitor monitoring
Go to usage instructions
We keep the neural network oiled by training it with new patents and search reports published weekly and validate its performance using real patent cases. The database covers about 75 million patent publications from all major countries and patent systems.
Book a demo
The first prior art search is on us. We hope you love it!