How Iconic went from disclosures to filed applications in days.

Iconic is a London studio that closed a $13M seed round to power the next generation of interactive storytelling experiences. Their first product, Digital Actor, is a technology that lets players speak naturally with AI-driven characters who improvise in real time while still landing the story the writer intended. Borja González León is Iconic’s Chief Scientist. He did a PhD in reinforcement learning and grounded language learning at Imperial College London, followed by research stints at Meta and Sony AI.

“What I was seeing was that game AI, and AI for games in general, was focused on redoing the things we have been doing for years, but just smarter and faster. I joined Iconic thinking about how we can bring new forms of entertainment through AI.”

— Borja González León, Chief Scientist, Iconic

Iconic’s public demonstration of their Digital Actor tech is Pressure Point. You’re on the bridge of a failing submarine, talking over the radio with James, an officer facing an impossible decision. He responds to what you actually say, so the lines are never the same twice, but the story still arrives where the creative director put it. Much of the field assumes language models can’t hold a script, and Pressure Point was built in part to show that they can.

A new mechanic like this is worth protecting early, ideally before competitors catch up, and certainly before a fundraise or the next public demo. Protecting it takes a patent process that moves at the same pace the company does.

Where the time actually goes.

A patent application lives or dies on inventor input. Only the research team knows how Digital Actor actually works—how the idea started, what they tried, what failed, what they eventually landed on. No one can supply that for them.

The difficulty is that an inventor usually can’t know which of those details a patent needs, because that’s a legal question as much as a technical one. A researcher given a blank page will write ten pages about whatever was hardest to build and one line about what’s actually novel. Turning that raw material into an application takes rounds of questions, and every round takes time from the one person who can answer them.

At a seed-stage company, that person’s time is the scarcest resource there is.

“Right now we are a small but growing team of full time researchers. If I remove my interns, the traditional patent process can quickly mean half of my team spending a week or two just on patents.”

— Borja González León

Iconic’s CEO, Andrew Bowell, spent a decade at Unity watching technology get built and defended, and he brought Fearn to the team. He already believed patent work mattered. What he needed was a patent process that ran at the pace his company planned in.

“Speed is the whole game at seed stage. A patent process that takes a quarter doesn’t protect you, the field has moved by the time you file. Fearn works at the speed we actually operate at.”

— Andrew Bowell, CEO, Iconic

What made it fast.

Fearn is a patent firm that builds its own software. Its platform, built specifically for this work, shows the inventor up front what the application needs to know, so drafting starts from a complete picture rather than a blank page. And because the practitioners writing the application work inside the same system the inventor does, the rounds of questions that remain move at software speed.

Fearn iterates the way patent counsel always has: draft, feedback, refine. The difference is turnaround.

“The back-and-forth is still there—Fearn asked questions, we gave feedback, the draft got sharper. The difference is each round took hours instead of a week. You never lose momentum.”

— Borja González León

Iconic supplied the technical substance, which only Iconic could supply. Fearn wrote the patent. The revision cycles between them ran in days.

Faster the second time.

The second application went faster still. González León knew by then what mattered and came in already targeting it, so the first draft landed much closer to what he wanted. Fewer revision cycles were needed, and the remaining ones ran at the same few-day cadence.

The speed didn’t come at the cost of coverage. Because the platform defined what the application needed before drafting began, the filings captured what was actually novel about Digital Actor.

Iconic now has multiple applications on file covering core aspects of the Digital Actor system, drafted on the research team’s schedule.

Who this is for?

González León is specific about where this fits: it’s a startup problem. Large labs have in-house patent teams and can absorb a process measured in months. Startups are racing competitors, disclosure deadlines, and their own next demo, and they usually file fewer applications than they should, later than they should.

“For a startup, the most precious thing is the time of each person. So anything that helps us streamline the process is very, very helpful.”

— Borja González León

Iconic set out to build a form of entertainment that didn’t exist before, which makes it hard to describe, since there’s no prior version to point at. That description still has to start with González León’s team—but now it starts from the right questions instead of a blank page.

“[Fearn] is doing something that is brand new, and I know how valuable this is. I would really like to keep using it. So far, we’re very happy.”

— Borja González León

If you’re sitting on research that needs protecting faster than a demo cycle, schedule a call.