USPTO Selects Instabase to Automate Patent Documents
The United States Patent and Trademark Office (USPTO) has successfully completed a pilot with Satsyil and Instabase’s automation platform to help automate signature extraction from inventor oaths for use with validation of micro entity certifications.
As discussed in the USPTO director's blog, the USPTO receives millions of patent applications and supporting documents each year. During the application process, inventors sometimes submit documents using alternative versions of their names.
This can limit the office’s ability to accurately certify the number of applications from a specific inventor and determine whether inventors are following application fee rules and regulations. Certifications submitted with errors often mean longer wait times for all applicants.
Prior to using Instabase’s technology, identifying patent application discrepancies required manually reviewing millions of documents to match names and signatures. Patent documents come in different formats and languages and can contain multiple inventors for each application. Signatures exist in various locations within a document, making signature to applicant name matching challenging.
Instabase trains and deploys state-of-the-art deep learning models to extract and match signatures and names from a near-infinite range of document formats and languages. The platform also expands on the USPTO’s efforts to incentivise positive innovation while addressing systematic abuse.
“As we explore new ways to bring America’s innovation to impact, Instabase allows us to quickly tap into more of the trapped data found in our unstructured patent applications with ease and accuracy. Reducing data barriers and openly preserving customer trust is a ‘signature’ accomplishment that we can all be proud of,” said Scott Beliveau, Chief of Enterprise Advanced Analytics at USPTO.
By integrating Instabase into the Satsyil workflow and infrastructure, the USPTO receives the following benefits:
- Data output from Instabase that allows for cross-correlation to find variations in names and signatures of the same applicant
- Alleviation of manual analysis to match names and signatures and accurately identify the number of submissions from a given applicant
- Proof of concept to apply similar Instabase extraction to other patent and trademark documents received by the USPTO
“For the first time in history, technology has come far enough to replace obsolete manual processes and truly tackle solving unstructured data,” said Erin Hawley, Vice President of Federal at Instabase.
“We are excited to work with a leader like USPTO who is showcasing the power of full-fledged automation for this critical use case.”
To learn more about USPTO and see the results of the recent data challenge, visit here.