The manual and laborious process of prior art search and review
Roshan Mansinghani serves as Assistant General Counsel at Unified Patents (“Unified”), a unique organization on a mission to improve patent quality by initiating targeted invalidity assertions against overbroad patents. With nearly two decades of patent litigation experience, Roshan is always on the lookout for high quality tools to support his practice and was an early adopter of NLPatent. In the early days, Roshan and his team used NLPatent intermittently to fill gaps in the art found by third-party search firms. They found NLPatent to be particularly effective at this task, especially when specific concepts or limitations were missing from the identified references by search firms.
In one such instance, involving a patent for sensors in automotive vehicles, Roshan used NLPatent to successfully uncover a key reference for a missing limitation, revealed within the first 5 search results! Recognizing the efficiency and accuracy of NLPatent, Roshan saw an opportunity to expand Unified's initial use case and leverage NLPatent to improve process efficiency for other operations.
The pros and cons of crowdsourcing prior art searches
Unified Patents is one of the few organizations taking a unique approach to support invalidity challenges through a crowdsourcing platform called Patroll. Patroll is a platform enabling prior art bounty contests, offering rewards of $2,000 to $5,000 to members of the public submitting relevant prior art with accompanying claim charts. Each contest runs for 60 days, after which an average of 10 submissions go through an internal and multi-layered review process by Unified. While generally effective, the review process takes considerable time (generally, about a month, at best), and occasionally they need to supplement the submissions with their own searches to meet quality standards, which is time-consuming and resource-intensive.
AI-based search outperforms crowdsourcing with precision and efficiency
Based on Roshan’s recent success with the platform, he wondered whether NLPatent could be used to find art for one of the Patroll contests. To his delight, NLPatent exceeded expectations. By simply inputting a patent number into the Patent Number Search feature, and adjusting a few filters, NLPatent quickly identified several highly relevant references on the first page of results. In less than 30 minutes, Roshan searched and analyzed the key references, resulting in several weeks of saved time and improvements in search quality.
“Oftentimes, we have an idea of what we’re looking for and while we have used other tools, the most recent version of NLPatent has proved to be quite effective in finding specific concepts.”
This was a step forward for Roshan’s team as they leveraged NLPatent to streamline operations in new facets of their organization, adding to their initial use case. NLPatent's ability to generate targeted, highly relevant prior art results instantly, without extensive input or modifications, make it an ideal tool for their fast-paced needs. The platform’s simplicity, speed, and functionality, allowing users to view descriptions, claims, and more in a single pane, are added bonuses rivaling even the most ubiquitous search tools. This efficiency significantly improved their triage workflow, saving time and resources for the Unified team to achieve their goals.
Peter Finnie | Partner, Potter Clarkson
How NLPatent impacts IP strategy, enabling patent attorneys to act as trusted advisors by providing immediate and actionable insights.
.svg)
.svg)