NLPatent: An AI-Powered Platform Revolutionizing Patent Data Access
We met with NLPatent Co-Founders Stephanie Curcio and James Stonehill to hear their founding story, learn more about their platform and the team making it happen, and welcome them as one of our Fund II portfolio companies.
Living in the golden age of innovation, it’s ironic that we rely on outdated technology to protect new ideas and inventions. When it comes to the patent process and defending intellectual property, it’s still done the old-fashioned way: accessing data by searching through rigid keywords and sifting through thousands of results.
Knowing there had to be a way to modernize the conventional process, Stephanie Curcio and James Stonehill founded NLPatent to revolutionize the patent domain.
“There are so many novel applications for AI to modernize the patent process, and that's what gets us really excited.”
Patent filings worldwide grew by 1.7% in 2022, with innovators submitting a whopping 3.46 million patent applications. With record-breaking figures like those continuously climbing, it quickly becomes clear why the decades-old patent process is problematic.
Consider this: you're trying to find the right keywords to express every single way your innovation has been described by anybody on earth, ever. You could either end up with a long list of results that capture inventions that may have nothing to do with your idea or a narrow subset of patent documents that are accurate and relevant but missing critical results because you haven’t considered every way to describe it.
Compounded with the hundreds of millions of patents already existing (and growing), searching the old-fashioned way is unsustainable—impossible, really.
Stephanie, an intellectual property lawyer specializing in patent prosecution and working with companies across different stages of the patent data process, experienced these inefficiencies firsthand. So, when introduced to modern natural language processing (NLP) technology by her original co-founder, this became an 'aha!' moment for Stephanie around the obvious use case for NLP in the patent process. Enter: the original proof of concept, and first version of NLPatent.
Then in 2021, Stephanie met James through one of the business’ early investors. With his software engineering background, specifically focused on machine learning-enabled products and services, James was the perfect fit as technical co-founder to help NLPatent reach its full potential, and the pair re-founded the business.
James and the team completely overhauled the platform, developed a new language model, rewrote every single line of code, and brought a successfully overhauled product to market—all while carefully nurturing the first wave of users. It was an ambitious feat that paid dividends: ahead of the ChatGPT boom, they bet big on building a large language model (LLM) to solve the patent problem, came out of beta in 2022, and haven’t looked back since.
“What we’ve created opens up and democratizes the patent process. Anybody—whether or not they have expertise in the nuance of patent language—will be able to access patent data for the first time ever.”
With its proprietary LLM, NLPatent can understand the nuanced language of patents. Rather than matching keywords, users describe their invention in full sentences, the AI system reads the entire text and distills the invention’s concept, and an intelligently ranked list of patent documents is provided based on actual similarity. The AI then learns from user input that steers it in the right direction to generate a refined list of results and plain-english analysis between the user's query and each relevant result. This is important as it bridges the gap between the complex language in the patent document and common parlance to ensure the user is guided to the correct portion of the document to review and consider its relevance.
NLPatent changes the way people interact with patent data, making the process much more effective and efficient for patent professionals and non-experts alike.
From here, the long-term plan is to build an end-to-end AI-first, IP intelligence patent management platform that seamlessly takes users from the idea stage, through the patent process, all the way to monetization and commercialization. AI will be applied to the patent portfolio to identify the greater picture, competitive intelligence, and all the bells and whistles associated with it.
With NLPatent's competitive edge of being built on modern techniques combined with their domain expertise in the patent and AI, NLP, and machine learning spaces, they are well positioned for this growth. The whole product is already architectured with an AI-first approach, and the team of twelve is small and nimble enough to continue leading this trend.
With the close of their seed round, the focus is on their go-to-market strategy and developing sales and marketing functions to support growth initiatives.
“We see ourselves at the intersection between AI and IP, bringing a modern AI-first approach to every step of the patent process.”
Stephanie, James, and the team have already surpassed many impressive milestones getting NLPatent ready for a larger market.
In the summer of 2023, they were one of 11 North American companies selected to participate in the Google for Startups Accelerator: Women Founders program. The 10-week virtual bootcamp provided high-quality mentorship opportunities and support for women in the tech startup ecosystem.
That same busy summer, NLPatent was selected as a top 3 finalist for the Collision PITCH competition. Selected from over 1,500 startups with representation from almost 30 industries and 76 countries, taking the stage was one for the books.
And the momentum of accolades has continued as Stephanie has been named a Global Leader in IP Strategy by IAM magazine three years running. Additionally, Stephanie is one of DMZ's 2024 Women of the Year and one of The Peak’s Emerging Leaders 2024 in the Legal category.
Recalling the complete platform rehaul that was a turning point in itself, pieces fell into place simultaneously with two other major life events: the pandemic and Stephanie’s pregnancy.
On the pandemic front, Stephanie and James initially met online thanks not only due to COVID-19 but also because they live in different countries: Stephanie in Toronto, Canada and James in London, England. The duo spent several months getting to know one another, a year working together virtually, and finally met in person in the fall of 2022.
And while on her side of the screen, Stephanie was preparing for motherhood and navigating the challenges that came along with it while building a business. Before James joined full-time, Stephanie shared that she was expecting and asked if he was sure he wanted to join the company. Not everyone would say yes in that situation, but James was in 100%, and now they’re making history with NLPatent.
“I had a baby in March of 2022 during the rebuilding phase, and it was great; it was sort of like the birth of my child and the rebirth of the business at the same time, which was a special thing.”
Check out a product walk through led by Stephanie in their Artificial Lawyer feature here.
Learn more about NLPatent at https://www.nlpatent.com.