Startup Mantra: Unlocking the potential of efficient IP management - Hindustan Times
close_game
close_game

Startup Mantra: Unlocking the potential of efficient IP management

BySalil Urunkar
Feb 10, 2024 05:02 AM IST

Simple way to search, review, analyse and collaborate over patents to address the needs of businesses and intellectual property assets

Pune: In the fast-paced world of intellectual property (IP) management, organisations often grapple with the daunting task of navigating vast repositories of patent data to extract meaningful insights. The challenge is further compounded by the complexities of competitive intelligence and the imperative to make strategic decisions swiftly and with confidence. Amidst this landscape, Manish Sinha-founded “PatSeer” emerged as the leading AI-based tool for patent intelligence and analysis.

PatSeer core team (from left) Prashant Mehta, Manish Sinha, Gargee Patankar and Tushar Surti. (HT)
PatSeer core team (from left) Prashant Mehta, Manish Sinha, Gargee Patankar and Tushar Surti. (HT)

“Patents are everywhere, from shoes, to farm equipment, medical, space, etc. Law firms, legal process outsourcing (LPOs), knowledge process outsourcing (KPOs), market research firms, innovative companies from different sectors like chemical, pharmaceutical, auto, electronic devices, universities, always need to extract intelligence from the text of different patents.

Now catch your favourite game on Crickit. Anytime Anywhere. Find out how

“Startups which innovate and then focus on scaling up initially rely on service providers, but once they reach a critical mass of 5 to 10 minimum patents, they start looking out for an in-house resource. At that stage, with a vision of continuous innovation for elevating product offerings, companies need a platform like PatSeer to do their job efficiently,” said Manish.

“PatSeer started as a search engine for patents, but since we had analytics background, our whole aim was to get technology-based insights by extracting intelligence from the text of patents online and provide all patent-related statistical analytics needed by companies. PatSeer makes it simple to search, review, analyse and collaborate over patents to address the pressing needs of businesses seeking to unlock the full potential of their intellectual property assets. By revolutionising the way organisations approach patent analysis and IP management, Patseer aims to alleviate the inherent complexities and empower enterprises to thrive in an increasingly competitive marketplace,” he said.

Struggle

An electronics and communication engineer from BITS Mesra, Ranchi, Manish worked in a multinational corporation near Delhi for a few years. He later quit his job to start his own company and came to Pune in 2004.

Manish said, “The core idea at that time was to process large volumes of text and extract insights, which would be applicable to any industry. For example, financial services sector has company reports, company filings, various numerical aspects are analysed by different portals, but there is not much on the relationships between company, its board, initiatives, and the way it is being used. We struggled for a couple of years doing different kinds of services. We also did data mining for bioinformatics. Then we got an opportunity to do some patents analytics.

“Patents are typically invention filings by companies. They disclose the details of the invention in return for a protection guarantee by the government. Patents are typically 20-to-30-page documents of an invention disclosure. They are all filed with the patent offices in different countries, wherever you need to protect it. So, we did a couple of projects in mining of these patents. And that is where we saw an opportunity to productise this service,” Manish said.

First product

Manish said, “Our first product was Patent Insight Pro, which was like a platform that sits on your desktop or a server, just like an MS Excel for patents. Users could bring in patents and the platform would give a wide range of tools to analyse it, along with statistical standard fields like when it was filed, how much, how many, what trends are there, etc. Trend and statistical analysis combined with text analysis gave the user a complete picture. At that time, we were selling in to around 15 countries, but we were relying on the patent content provided by third parties. We did not hold the content and that was putting a limit to where we were. We realised that we could only do so much with an analytical tool. We had to go into becoming a worldwide patent aggregator and a publisher ourselves. This required a lot of data aggregation from multiple patent offices, purchasing it from third party providers, putting in teams to process and normalise, clean up all the information was a very big project. And that is when I personally also invested, and PatSeer was launched by the end of 2012.”

Patent search engine

Manish said, “When we started off, PatSeer was essentially a worldwide search engine for patents. The basic workflow was you search, you view, browse the results, export the results, set up alerts on areas of your interest, etc. To identify our market, we thought about who needs a product like this. The answer was simple. Companies that are in continuous innovation, who file patents periodically, they need to not just periodically monitor the competition, but also monitor the trends in their space. If such companies are getting into lateral spaces of their products, then they need to evaluate what kinds of patents exist out there, who are the threats, what are the challenges, who could pose a challenge to them, etc. A lot of this information for innovative companies, comes from patents. And tools like PatSeer is what it is used for.”

“In India, for patents, the traditional sectors have been pharma and biotech and a bit of electronics. Now automotive industry, especially electric vehicle patent market, is expanding. Hence, initially for PatSeer, the market was not just India but also the US and Europe,” he said.

Product development

Further editions of PatSeer, with different levels of subscription, were launched in four years. Explaining it, Manish said, “The base edition does the basic job of search, alerts, exports, etc. The next edition combines all features of the previous one, but is also for organisations who want to collaborate between IP and R&D teams. They can set up internal workflows and share patents via the platform itself, all of those needed by corporate, because nobody simply just wants to always export to an Excel, or a PDF, or a Word, and then share the doc. Why not give the share via the interface itself, the person receiving can, add their value, add their inputs in form of, rating, marking records, or comments, annotations, whatever, via the platform itself, and thereby you can bring that, whatever project you have had to a conclusion, or the decision-making process. The third is the analytical edition, in which one can have, competitive intelligence workflows, monitoring of competitors, and additional statistical insights needed for.”

Crafting IP strategy

PatSeer helps companies to craft their own IP strategy. It provides insights of dominant market challengers, companies holding the most valuable portfolios, acquisitions, litigations related to patents, etc.

Manish said, “Mid or large size corporations, who have a portfolio of 50 plus patents actively craft a product and IP strategy. Because IP strategy is needed at every stage of the product development. For example, a new idea needs IP evaluation, and by the time product is out, the IP docket needs to be ready. This helps in supporting the product commercialisation strategy without hitting any roadblock in the market, or anyone posing a challenge to the product or preventing anyone from copying the product. PatSeer provides all such insights needed by the analysts or senior product technologists, managers in-house or at any service providers.”

AI opportunity

PatSeer started incorporating artificial intelligence (AI) in its platform since 2017, when it was not generative AI or the Large Language Models (LLM). Now PatSeer deploys a combination of LLM to provide an AI-based search engine for patents. “Patent searching is not as simple as just Google searching for patents,” Manish points out.

“We had the rules-based semantic search engine which is now combined with the latest AI stack models. Users write complex Boolean search queries which can be a page long to search deep within the technology. This we call as an expert search. AI search is one where a user comes up with an idea, describes it in a paragraph, and PatSeer will show them exactly, which are the closest links to it or analyse uniqueness of the idea. PatSeer also allows organisations to give access directly to the R&D persons so that they can look up their ideas, and check for its novelty in the early stage itself. This increases the market scope of a new product,” said Manish.

“AI has expanded the market for PatSeer. We create a training data set on patents, from existing patents and then we feed that into an already trained language model. It understands the language of patent claims, the legal portion of patent describing the invention, etc. This training also allows it to understand the intent of that complex representation. Thus, AI allows us to answer low, medium level cognitive queries and leave the deeper analysis and analysis that needs a far deeper assessment to the experts. Previously users were able to get level-one insights, but now they can get level-two and three insights also. Business managers can get these insights because systems are becoming simpler, and intuitive and more insightful with lesser number of steps using AI. This is widening the scope of AI as IP is eventually being directly taken to the decision-making table,” Manish said.

Growth

Within three years of operations, PatSeer reached around 2,000 users from the R&D, market research, knowledge service, legal service providers segment. With current strength of 90 team members, PatSeer aims to expand globally especially in countries like Japan, China and Korea.

Manish said, “Developed countries are very aware of IP. In India, biotech and pharma has always been IP savvy, but we are seeing increased overall awareness in other sectors too. We have grown strongly in India. We have about 300 legal service providers and the Indian Patent Office using our platform. PatSeer now plans to provide predictive insights to users using AI. Generative AI would help us automate on certain legal aspects like drafting a patent application, replying to certain enduring IP prosecutions, etc. We are also looking for combining AI and Blockchain for trade secret management, copyright management. AI can also recognise patterns within images and hence design and image searches will be our focus area.”

SHARE THIS ARTICLE ON
SHARE
Story Saved
Live Score
OPEN APP
Saved Articles
Following
My Reads
Sign out
New Delhi 0C
Monday, June 24, 2024
Start 14 Days Free Trial Subscribe Now
Follow Us On