More often than not, innovators become entrepreneurs to solve a problem or tackle a challenge that they themselves, have seen or lived. When Stephanie Curcio, CEO and Co-Founder of NLPatent, noticed inefficiencies in her own Intellectual Property practice, Curcio knew that it was time for a change, starting with the patent search process. “When I was practicing, I used to manually look for patents which meant searching through hundreds of millions of patents using keywords to see if our client’s idea has been patented before. Patents are very dense, complex, and technical documents, making them notoriously difficult to read and understand, let alone search” shares Curcio. For a patent search to be accurate, it is important for patent professionals to craft the perfect query with the right keyword strategy, which involves thinking of every single way a technology has been described since the beginning of time. This is, understandably, extremely difficult in the information age. On a mission to modernize traditional patent and innovation processes, Curcio’s venture, NLPatent, started by revolutionizing patent search using the power of AI.
A multi-value proposition solution, NLPatent leverages modern Natural Language Processing (NLP) and Machine Learning (ML) techniques to support IP and R&D professionals in finding more accurate patents related to a given invention in an affordable and efficient manner. Purposefully building their ML models to understand patents and the language of innovation, their proprietary AI solution thoroughly understands the invention by accepting an entire paragraph of information in natural language, rather than a Boolean search string that blindly matches keywords. The output is a set of patent results that are conceptually similar to the invention description, which can be further refined based on user input (i.e., “liking” or “disliking” results). “What we have seen is that our clients can do a search way faster and more accurately than ever before – minutes vs. hours. They are not generating hundreds or thousands of irrelevant results; they are getting a targeted list – one that often contains patents that traditional techniques would miss,” says Curcio.
NLPatent’s platform was originally launched in 2019; this innovative venture has a portfolio of clients including reputable law firms, Fortune 100 companies, and distinguished household names. With a fully remote team, the company took a major turn in 2021 when Curcio’s team revamped their underlying technology to adopt a proprietary fine-tuned language model purpose-built to understand patents. Reflecting on the process, Curcio shares, “We realized that a transformer-based, large language model was necessary to deliver the greatest value to our clients. This was the right decision for our business. The model we developed works extraordinarily well in addressing the problem of patent search and provides highly accurate results, outperforming human experts.” With experience and knowledge in both machine learning and patents, Curcio and her Co-Founder, James Stonehill, brought technical experts, including highly skilled data scientists, on board to help them create a cutting-edge solution that adopts the latest AI capabilities, providing an industry leading solution to their clients. “We are very serious about the technical strength of our product; not just claiming that we use “AI”, which has become an over-used buzzword, but actually building a real, innovative, and proprietary technology that is leaps and bounds better than an off-the-shelf language model,” says Curcio. An industry agnostic solution, NLPatent has seen various use cases across the patent and innovation lifecycle, even at the threatened “end” of a patent’s life during litigation. Often, companies that are sued for patent infringement would use NLPatent to invalidate the patent being asserted against them. NLPatent shines in these scenarios because it approaches the patent search in an entirely different way, generating highly relevant results that were previously missed, rendering the patent invalid and therefore unenforceable.
Celebrating World Intellectual Property Day today, we are celebrating inventors and creators who are driving technological innovation from around the world. With this year’s theme of Women and IP: Accelerating innovation and creativity for this World IP Day, we asked Curcio about her thoughts on women representation and gender disparities in this field. “Becoming a mom has opened my eyes to gender disparities; I have started noticing natural but apparent differences in gender roles a lot more now. There are more responsibilities and expectations on women, which are difficult to balance. The numbers speak for themselves – we see a lot fewer female innovators and tech founders. My only message to all the women out there wanting to pursue these fields is to surround yourself with supporting, encouraging individuals. Go for it – you’ve got this!” says Curcio.
Looking to the future, NLPatent is full steam ahead on further developing their technology to modernize all aspects of the innovation lifecycle beyond patent search. “We’re seeing that companies can use the same underlying semantic understanding of innovation to better understand their own patent portfolios, maximize the ROI of their R&D efforts, and make better informed decisions on innovation strategy,” says Curcio. The company has also seen a rise in the use of their software by non-patent professionals to devise an intelligent and well-researched competitor analysis, identifying white spaces in their field of technology. “Our traction is testament to the amount of excitement and energy around the application of Large Language Models in the patent sphere, and we are excited to be along for the ride! We have a strong foothold in the market and are in a great position to take this to the moon.”
Working closely with our Manager of Startup Advisory and Venture Growth, Andrew Leest, NLPatent is helping their clients maximize efficiency and create well-researched IP strategies. Leest shares, “We are constantly working with client companies to assess the IP landscape around their innovations and create a strategy to build a strong portfolio over time. A comprehensive patent search is one of the earliest points in that process that can give a company freedom to operate and comfortability investing further in their innovation space. NLPatent is allowing this first hurdle to be cleared in a fraction of the time that a manual search process would take and set the entire strategy up for success in the future.”
To learn more about NLPatent and their solution, visit https://www.nlpatent.com/