Skip to content

Prior Art Search for Everyone

A collaborative initiative to build a common AI-based prior-art search tool.

An initiative by AT&T


PQAI stands for Patent Quality through Artificial Intelligence.
Through this project, our vision is to make the patent process more transparent for all involved, significantly grow the number of inventors, thus accelerating the pace of innovation, and simultaneously improving the overall quality of patents.


Focus on Precision

Very few researchers have time to read hundreds and thousands of results. The long-term goal of PQAI is to bring the best prior-art in the top-10 results.

Big on Privacy​

PQAI never logs any of your search queries or results unless you explicitly tell it to save them for later reference after creating an account.

Open Source Algorithm

Building the next-generation prior-art search search system is no small feat. We believe that it would be much more efficient for the industry to collaboratively build a common open source tool.

Ease of Use​

Prior-art search need not be limited to professional searchers. PQAI aims to provide a simple interface that allows anyone to check the novelty of their ideas.

PQAI Roadmap

The idea of PQAI was conceived
November 2019
First prototype search engine covering wireless technology
November 2019
February 2020
Coverage expanded to all technologies covered by US patents (except Chemical and Biotech)
February 2020
June 2020
Released a new version of 102 (anticipatory) prior art engine
June 2020
November 2020
Combinational Search (103 engine) Beta version
November 2020

PQAI will benefit

Power of PQAI for Inventors


Power of PQAI for Patent Examiners

Patent Examiners

Power of PQAI for Law Firms

Law Firms

Power of PQAI for Enterprises


Initiative Started By

PQAI Contributors

Algorithm Contributed By
Search App Managed By

Frequently Asked Questions

 At PQAI, we believe that 10 results are enough for inventors to help him determine the novelty of their inventions. We are continuously training our engine to keep increasing the relevance quotient for the results.