.png)
Picture this: A patent attorney sits down to assess patentability of an AI-based system for searching patent documents. But instead of performing tedious boolean searches or outsourcing to a third party, they turn to the very software they’re trying to patent - to help patent itself. If that sounds like inception, you’re not alone!
But “Inception” is exactly what happened when the founders of NLPatent turned to their friends (and clients!) at Potter Clarkson LLP to secure intellectual property protection for NLPatent’s proprietary system of large language models (LLMs), trained specifically to speak the language of patents.
The firm, known for its strategic IP advisory work with startups and scaleups, assigned Angus Gledhill to the matter - a seasoned patent attorney with deep expertise in AI. Luckily for us, Angus was no stranger to the platform.
Angus and his colleagues have been using NLPatent for over a year to streamline natural language search, eliminate manual keyword guesswork, and level-up their early-stage strategy work. By this point, NLPatent had become a go-to resource at the firm. From uncovering knockout references in seconds to highlighting the most relevant snippets with Relevance Analysis, the platform had consistently proven its worth - saving time, increasing accuracy, and transforming early-stage search into something… well, bearable.
So when Angus was tasked with assessing the novelty of NLPatent’s own system, he did what he’d done so many times before: he logged into NLPatent.
Anyone who has worked on AI-related patents knows the struggle. The underlying tech is often complex, nuanced, and (let’s be honest) hard to articulate in crisp claim language. That’s where NLPatent shines. It’s particularly adept at handling language that’s difficult to distill into strings of keywords, making it uniquely well suited for evaluating AI-related inventions.
As is typical in his workflow, Angus began by chatting with the inventors (in this case, NLPatent’s CTO and a Senior Applied Scientist), and drafting a quick pseudo-claim to zero in on what seemed to be the novel element. Then came the magic moment: he ran that concept through NLPatent.
Refinements and iterations followed, focused tweaks that revealed what else was out there, what had been done before, and more importantly, what hadn’t. The results confirmed it: while other AI search tools existed, none captured NLPatent’s unique approach. The invention appeared both novel and non-obvious. It was go time.
Angus didn’t stop there. He kept NLPatent open throughout the drafting process, treating it like a co-pilot. Claim by claim, he refined scope, re-ran searches, and adjusted based on the most relevant prior art the system uncovered - often in seconds. The feedback loop was immediate and intelligent.
Had he followed the traditional path, this might’ve looked very different. A patentability search (if conducted at all) would have been outsourced. A static report would land in his inbox a week or two later. Whatever that report contained would shape the claims, whether it was complete or not. There’d be no real-time insights, no ability to pivot mid-draft. But by being exposed to the art first-hand using NLPatent, Angus gained so many intangible benefits, providing unparalleled benefit to his client (us!), doing it all himself - refining, rerunning, iterating - on the spot.
The result was a thoughtful, well-supported application for a platform that’s already transforming how the industry does patent research. Talk about full circle!
As Angus put it:
“For an IP lawyer, it’s one thing to use a patent research software to assist clients. It’s another to use that same software while drafting claims for its own patent application.”
We couldn’t have said it better. This story is one of our favourites - not just because it’s ironic and kind of hilarious - but because it showcases what’s possible when the right technology is in the right hands. NLPatent helped patent NLPatent. And in doing so, it proved its value in the most poetic way possible.
Sign up to receive alerts when we publish more blog posts about how NLPatent has brought success to the field of IP and AI.
.svg)
.png)
.svg){kind=small}
