Topics
previous
AI
Amazon
Image Credits:Peter Dazeley(opens in a new window)/ Getty Images
Apps
Biotech & Health
Climate
Image Credits:Peter Dazeley(opens in a new window)/ Getty Images
Cloud Computing
Commerce
Crypto
Enterprise
EVs
Fintech
Fundraising
Gadgets
gage
Government & Policy
Hardware
Layoffs
Media & Entertainment
Meta
Microsoft
concealment
Robotics
Security
Social
quad
startup
TikTok
Transportation
Venture
More from TechCrunch
Events
Startup Battlefield
StrictlyVC
Podcasts
Videos
Partner Content
TechCrunch Brand Studio
Crunchboard
Contact Us
Founded in 2007 to serve IP and R&D professional do their jobs more expeditiously , Singapore - basedPatsnap‘s first product was a global letters patent lookup database . Now it ’s build out its suite of AI products with the launch of its AI supporter CoPilot .
CoPilot lets user lookup letters patent and non - apparent literature more speedily , which the caller says will make IP and R&D workflows faster across Patsnap ’s full mathematical product rooms . Its proprietary LLM also link up back to sources , include diary and patents , and provides references .
The inauguration says it has spend million of dollars and has a squad of more than 50 technologist dedicated to its AI capability . Patsnap CEO and co - founder Jeffrey Tiong tells TechCrunch that Patsnap live to remove friction in the innovation cognitive operation for its customer , both within IP and R&D teams , and between them . IP analysts and attorneys need to run prior graphics ( selective information publicly disclosed before a patent was filed ) and freedom to function ( or the ability to utilize a intersection or service without facing effectual issues ) research to make up one’s mind where their companies should clothe meter and money , and Patsnap ’s intersection are design to make the process more smooth .
For example , Patsnap ’s Analytics product include over 180 million patents and over 130 million small-arm of lit from 170 jurisdiction . informatics teams can utilise Patsnap ’s AI tool to take apart their markets and protect invention at scale , state Tiong .
What CoPilot does is build further onto Patsnap ’s AI ware . It enable IP and R&D squad to bump what they need more apace within patents , non - manifest literature and expert news show . Some examples Tiong give include getting automatic summaries of patent claims , see about a technology and bewilder radio link to a company ’s patents . It also answer specific problems . For instance , if a R&D professional want to know more about improving battery energy density , CoPilot can translate patents and find relevant lit .
CoPilot also aid R&D and IP teams in four other ways : preserve up - to - particular date with rapidly changing sector , supply content depth psychology to help guide strategic patents and research , extract primal details from specific patent and literature , and AI security because its model is n’t educate on customer data . Tiong says that since CoPilot has its own proprietary LLM , client data does n’t call for to leave behind Patsnap ’s firewall , so it is n’t passed to outside networks .
Tiong adds that Patsnap ’s LLM was trained on information from patents , academic newspaper publisher , expert theme and recent company newsworthiness , including about merger and acquisition . It also incorporates information gloss by IP experts and Patsnap ’s products .
Join us at TechCrunch Sessions: AI
Exhibit at TechCrunch Sessions: AI
Its LLM learns generic information , specialized data and aline data in three stages : pre - train , place pre - train and ego - training fine - tuning , so it has specialized accuracy in patent and non - plain data . commingle with recovery - augment generation , Tiong say that compared to GPT3.5 , CoPilot ’s theoretical account is better at performing in - depth depth psychology while being less potential to hallucinate answers .
Patsnap has raised $ 350 million from investors like SoftBank and Tencent . It has a sum of more than 1,200 employees , with 12,000 customers . Its customer come from upright , including life sciences , self-propelling , consumer goods , technology , manufacturing , applied science and sound .
