HQ: Helsinki, Finland
Invested in: Q2/2018
Founders: Sakari Arvela, Juho Kallio
Keywords: IPR, patent search, natural language processing, AI
How AI is Making the Patent Search Process Better
The taste of paper springs to mind when hearing the word patent. For a reason. During the patent process, from searching and drafting patent applications to review these documents, millions of hours are spent (and wasted) combing through bottomless databases and reading boring technical descriptions.
Any task requiring a lot of time that uses a lot of data is poised for improvement through AI (artificial intelligence). IPRally, a Finnish startup, aims to do just this: Improve the patent search process using AI.
IPRally’s search function isn’t just a complex keyword or similarity search. Instead, it relies on a data model which represents the way patent attorneys really think, the prosecution history of patent applications and a neural network that learns the key logic of patenting. In essence, their AI has been trained by expert patent attorneys and examiners.
In order to perform a search this complex and detailed, the team utilizes a Tree-LSTM-based deep learning solution. The company hopes to be able to continue improving the model’s speed and scale going forward for harnessing real-world patent examination data to serve the IPR field.
They have an intuitive user interface to help target the search better, which results in more meaningful results. Users don’t need to have prior patent experience. Any engineer can quickly use this tech. Instead of wasting time with manual analysis and vague searches, with IPRally users can find patents that really matter.
As with many business ideas, IPRally was founded after asking questions that couldn’t be answered at the time. Sakari Arvela, CEO and co-founder, is a European Patent Attorney by trade and saw the need for a better way to handle the patent process. He partnered with Juho Kallio, an experienced software developer who is now CTO, to found IPRally to address these unanswered questions. They’ve rounded out their team with artificial intelligence experts.
Venture capital financing will enable the team to continue building out their technology and to focus on other developments that influence the future of patents and IPR.