Democratizing the patent industry    | TechAlliance of Southwestern Ontario

Patent law is the legal system regulating the granting of patents for inventions. The process involves lawyers who, along with other tasks, have to make sure that the proposed “idea” has not already been patented, which can be complicated, time-consuming and costly. But that was before NLPatent was launched. 

A Good Idea Takes Shape 

Stephanie Curcio is a lawyer who understands patent law and its many complexities. So, when a colleague with a background in artificial intelligence (AI) came up to her with an idea and prototype that would revolutionize the system, Curcio jumped on board.  

The decision to leave her job on Bay Street wasn’t easy, but Curcio was ready to leap into democratizing a patent search for innovators because it had so much promise.  

The Idea Becomes a Business 

The owners of NLPatent realized early on that they would have to constantly evolve in the software space. Fortunately, potential clients seemed to understand that the process for patenting intellectual property (IP) could be improved, so the demand was there. NLPatent’s platform 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. 

After some initial traction on the first version of the product, Curcio began collaborating with a new partner, James Stonehill, and the pair launched the second iteration of the NLPatent platform. It was powered by a significantly updated large language model, and while it was a gamble given that the new proprietary model took a lot of effort and money to build, Curcio and Stonehill were ready to put their heart and soul into it. 

Streamlining the Search 

Patent language is unique. It’s complicated and diverse. As such, Curcio and her team had to create a sophisticated large-language AI model to not only understand the nuances but also to interpret and translate them. As such, they had to “train” the AI model using patent data because a generic language model wouldn’t work; patent documents are not written in plain English. 

The AI model was not only trained to interpret patent language but also to simplify it as patents are incredibly complex and can be 100+ pages long! . This length includes substantial boilerplate and background information outside of the scope of the patent’s innovative concept. Employing this proprietary methodology allowed for concept-based searching rather than keyword-matching or technology classification, which was a breakthrough. With over 120 million patents to sort through, the model looks for meaningful representations to determine similarity. And it does this instantly! 

TechAlliance Steps In  

NLPatent’s continuing evolution required support, and that support came in the form of TechAlliance. Curcio was introduced to the TechAlliance team in the spring of 2023, at a time when the company had demonstrated traction but also had a lot of moving parts that needed capital. NLPatent was encouraged by TechAlliance to complete the i.d.e.a Fund ^TM application.   

The i.d.e.a. Fund ^TM is supported by a $10-million Government of Canada investment, through the Federal Economic Development Agency for Southern Ontario (FedDev Ontario), led by Boundless Accelerator (formerly Innovation Guelph) in partnership with five Regional Innovation Centres (RICs): WeTechAlliance, Haltech, Innovate Niagara, Innovation Factory and TechAlliance of Southwestern Ontario. The program helps high-potential ventures access the tools they need to succeed by providing financial and business advisory supports to develop or redesign products, services, processes and technologies that reduce impacts on the environment and create made-in-Canada solutions. 

Timely Funding 

The funding came at the perfect time. NLPatent had enjoyed tremendous, accelerated growth and had created several jobs, but Curcio and her team were preparing to raise a seed round to fund travel, develop the business, and launch a further optimized language model.  

Before doing that, they had to complete thorough testing and benchmarking, which is expensive. With the funding coinciding with the scaling up of the company, NLPatent was able to subsequently double its revenue in a matter of six months.  

TechAlliance’s Support 

In addition to linking NLPatent with crucial funding, TechAlliance provided mentorship support. Curcio and her team worked closely with Venture Growth Advisor, Dr. Martin Smith, who was always willing to introduce Curcio to key contacts. Those connections have helped NLPatent achieve its goals, including closing a Seed Round of funding totalling USD $1.5 million.  

Where to Next? 
 
Curcio and her team, which has grown from five to 13 employees, are working on a new product they hope to launch commercially in Q3 or Q4 of 2024. The new software will leverage patent data to unearth valuable business intelligence that was previously labour intense and extremely time-consuming to uncover. The team is still developing the new product but did share that it will be more forward-looking, using an existing patent portfolio to track and monitor competitors (among other things). It will also assist in detecting infringement and acquisition targets, and help companies value their IP.  

“There is so much room for improvement in the patent space, and Stephanie and her team have shown that they are the company to make these strides and provide significant help to those looking to patent their innovative ideas,” said Andrew Leest, Manager of Venture Growth. 

Winners of the 2024 TechAlliance Limitless People’s Choice Award, NLPatent is well on the way to achieving its goal of empowering entrepreneurs in the vast IP landscape.  

To learn more about NL Patent, visit htttps://www.nlpatent.com
To learn more about the i.d.e.a Fund^TM, go to https://www.techalliance.ca/i-d-e-a-fund/ 
To learn more about FedDev Ontario, go to https://feddev-ontario.canada.ca/en