The IFI Interview: Jeff Roy, Co-Founder and CEO of FluidityIQ

Jeff Roy remembers the information industry before the internet, when everything started to change in a “slow-roll kind of way.” Storefronts popped up. Users started to see data differently. The way companies operated their businesses and engaged with customers transformed.

Today, Roy is seeing the sparks of a similar transformation—only this time, it’s about AI. “I recognized that AI technology had evolved to the point that it was going to change the way people interacted with information going forward,” says Roy in a recent discussion with IFI CLAIMS. “The emergence of AI represents the same kind of inflection point.”

How will that affect the IP industry? As it currently stands, Roy sees traditional players as operating in a mostly transactional way, especially with regards to offerings like patentability or freedom to operate searches. Going forward, he thinks delivering relevant patent results is no longer the end game, because innovation is not transactional; it’s becoming about providing a richer level of intelligence, swiftly. In other words, explaining the “why” and “how” along with the patent search results.

That’s the opportunity that led Roy and his partners to start FluidityIQ, a company that supports discovery and innovation for research pros. IFI CLAIMS sat down with Roy recently to chat about the company and where he sees it heading. He talked about the advantages of building the platform from the ground up with native AI, innovation as a life cycle and not a point in time, the nature of data in patents, and why he chose IFI CLAIMS to support the patent portion of the platform. The interview has been edited for clarity and length.

IFI CLAIMS: Tell us about your background. How did you get into the patent intelligence business?

Roy: I’ve spent over 30 years in the information industry. As for how I got into intellectual property and patents specifically–I was part of the management team that built Clarivate Analytics after its carve out from Thomson Reuters in 2016. I initially joined to run the CompuMark business, focused on trademark research and protection. Over time I went on to lead the full IP group after Clarivate went public, overseeing patent standards and all the other IP-related areas.

IFI CLAIMS: What led you to start FluidityIQ?

Roy: I started FluidityIQ in part because I saw an opportunity to do things differently. The IP industry is very transactional by nature. When it comes to patents, most solutions support what I would call maintenance activities—things like annuity payments. On the research side, it’s the same story: most of the solutions are focused on delivering a search as a one-off transaction, whether it’s a patentability search or a freedom to operate search. That never really made much sense to me because the data within and across patents holds valuable innovation intelligence–something far too important to treat like a checklist item. Innovators need to be thinking in terms of a life cycle, not just isolated tasks. That’s where FluidityIQ comes in: we’re here to help change the way innovation is unlocked and managed–both in traditional and novel use cases.

IFI CLAIMS: FluidityIQ uses native AI. Talk about what that means and what it brings to the process that maybe your clients wouldn’t get from a company that isn’t native AI.

Roy: A lot of the companies—not just in the IP industry, but across industries—are in a hurry to bolt a large language model, such as ChatGPT, onto their legacy application. And while that can improve productivity and make their tools a little easier to use, it doesn’t fully unlock all of the benefits that AI can offer. LLMs do some amazing things, but they don’t actually reason yet; they just do a great job of creating the illusion of it.

When we talk about Native AI, we mean something different. FluidityIQ was built from the ground up with a semantic search engine that finds relevant search results based on the underlying meaning of patent claims–not by relying on old-school Boolean operators or key word matches. A lot of players are layering LLMs on top of traditional search to simplify the experience. With FluidityIQ, you don’t have to work around old models. Native AI gives us, and our users, far more flexibility and power.

IFI CLAIMS: Take us through exactly what FluidityIQ does. What problems is it solving for your clients?

Roy: Our current use cases focus on traditional IP tasks, like freedom to operate. If I have a patent and want to enter a new market, I need to make sure that my patent doesn’t infringe on anybody else’s in that market. Or as an innovator, I may conduct a patentability search to see if my new product is patentable before investing further. What FluidityIQ does is make it easy to compare the features of your specific innovation to a universe of 160 million patents that have been filed around the world. What differentiates us from other solutions is being native AI. Instead of just returning search results, we explain why a patent is relevant to your particular use case—whether it’s patentability or freedom to operate. This approach significantly reduces the cost and improves the quality of patent research in these traditional areas.

Looking ahead, our long-term vision is to expand beyond traditional IP research to help diverse markets make better data-driven decisions. We believe that innovation data held within the universe of patent documents can create valuable intelligence for players in different sectors looking to invest in a particular technology and/or reduce risk–maybe helping banks reduce risk tied to collateral, for example.

For now, we’re focused on lowering the cost and improving the quality of the kind of research that traditionally happens in the IP industry, whether it’s for legal teams or R&D.

IFI CLAIMS: What problems did you see in IP that caused you to want to start FluidityIQ?

Roy: There are two key reasons. First, as I’ve already mentioned, the IP industry tends to be transactional, with a focus on one-off projects. We wanted to build a solution that helps people navigate the entire innovation lifecycle within their technology, not just address a point-in-time need for a research project around the patentability of their innovation.

The second reason relates to timing. I recognized that AI technology had reached a tipping point where it would transform the way people interact with information. Having spent 30 years in the information industry, I remember a time before the internet. I also vividly remember how the whole field of business operations changed when it arrived. It changed in a slow-roll kind of way. At first, people put up storefronts–nothing groundbreaking on its own, but without that step, we wouldn’t have e-commerce and online trading platforms today. I could see a change in the way people related to information. AI is at a similar inflection point in the information industry. People are quickly adopting tools like ChatGPT to find efficiencies, but the real shift is just beginning.

We’re at the tip of the iceberg. Simply conducting a patent search and getting relevant results back isn’t enough anymore. People want fast, contextual intelligence that says, “These patents are relevant to you and here is why.” That’s a whole new way of interacting with information. The capabilities now available allow for a much more dynamic, interactive experience than ever before. The combination of IP’s transactional nature creating an opportunity for a better solution and AI’s evolution in how we engage with information drove us to create FluidityIQ.

IFI CLAIMS: You talked about the transformational nature of the internet. How do you think AI will transform the way patent data is used in the future?

Roy: It’s going to speed up innovation. If the kinds of work that used to be transactional suddenly become embedded into an innovation life cycle, it will streamline innovation by making things like invention harvesting easier. It will also lower costs, but not in the ways that many expect. People fear that efficiency gains are going to cause people to lose their jobs. I believe it will transform roles instead. And in many cases, it will act as what I like to call a “force multiplier.”

IFI CLAIMS: Tell us about your platform.

Roy: Traditional patent search involves creating a series of Boolean operators or key words in order to identify patents that are relevant to a use case. In contrast, we’ve built the product around three core pillars that reflect our commitment to a holistic, rather than transactional, model. The first is search, which uses AI to enhance results, moving beyond Boolean operators.

The second pillar is curation, which harks back to the notion that simply returning search results is insufficient. We add a layer of intelligence to those search results, explaining why and how certain patents are more relevant than others.

The third and last pillar is persistence. Traditional transactional IP artifacts–a static patentability report for example–represent just a snapshot in time. This is fundamentally incompatible with the very dynamic and ongoing nature of the innovation lifecycle. We believe that a research initiative like a patentability search shouldn’t end after the initial query, but evolve as new data becomes available around the world—such as updates from sources like IFI. And we should be able to push that intelligence to our customers automatically. Our goal is to make it easier to understand the relevance of information and ensure that it forms a valuable knowledge base to advance research.

IFI CLAIMS: Is there anything else important to note about FluidityIQ?

Roy: I am far more interested in the data within patents than the patents themselves. I think there is significant opportunity for us to create useful intelligence and decision support tools by examining, for example, how patents in a particular innovation area relate to each other, or how companies are connected through the innovations they’re pursuing. To me, that’s more compelling than the patent document itself.

On a somewhat related note, we built FluidityIQ around the concept of a central personal library that allows users to combine different types of data in meaningful ways. It’s not just about patent data, it’s about blending that research with internal data, other research, and market documentation. Then we allow users to interact with this combined data using natural language.

IFI CLAIMS: What advice do you have for companies and IP professionals looking to modernize their patent research process?

Roy: Be open to new ways of approaching your work. It’s natural to worry that the availability of new tools might replace jobs, but it’s more important to understand how these tools operate and how they can benefit you.

I’d also advise being careful, intentional, and curious about what tools you’re buying and how you’re using them. The thing is that “AI-based” can ultimately pertain to anything from decades old machine learning algorithms to truly native AI platforms. It’s critical to understand how AI is being applied, what it does well, what it doesn’t do, and what the unique value proposition is in each case.

IFI CLAIMS: Why did you decide to go with IFI as your patent data provider?

Roy: We love how easy IFI is to work with. During our startup phase, when we needed to onboard large volumes of data quickly, IFI made it simple. We didn’t have to spend time preprocessing or cleaning data. We could load it and immediately focus on building our product, which is exactly where we wanted to direct our energy.