Wren reads your brief and maps the landscape - the markets where the core use case fits, the single best beachhead, and OTHER USES the same mechanic could serve in different problem domains. (Terry built SLR Studio for manuscripts; the tech turned out to power pharmacovigilance and patent monitoring. That pattern is common. Wren finds it on purpose.)
Wren reads your brief and returns: 3-5 market segments where the core use case fits (each with density label and a 30-day signal to measure), the single best beachhead with reasoning, 3-5 OTHER USES the same mechanic could serve in different problem domains, 2-3 adjacent product pivots, a white-space map (with opportunity / caution / graveyard verdicts), 5-8 specific search hooks you should run yourself to verify, and the one move Wren would warn against.
Wren reads your brief, identifies the 3-4 patent-search queries that would surface the relevant prior art, runs each one live against Google Patents, and produces: a prior-art landscape (5-10 patents with relevance and overlap labels), a patentability assessment (which claims could survive, which are likely DOA, where the white space is), 3-5 suggested CPC classification codes, and three next-step paths to file (provisional self-file via USPTO TEAS Plus, full patent attorney engagement, or USPTO Pro Bono Program). Takes 60-90 seconds. Not legal advice - bring the output to a registered patent attorney before filing.
Eight rules that separate a real landscape read from a Google search. The generator above follows these too.
Founders search to confirm their idea is unique. Wren searches to find what is already there. The first answer is almost always more than expected. The gap between expected and actual is the brief's first weakness.
Empty quadrants on the map are sometimes opportunity. They are more often graveyards where founders have already died trying. Find the graveyard before celebrating the empty space.
No filed patent in your space means one of three things: the IP exists as trade secret, the space is too small to bother protecting, or someone tried and the application was abandoned. None of those is the founder's friend by default.
Most ideas can serve more than one market. Wren names 3 to 5 because the second-best market often has the cleaner first signal. Do not assume the first market you imagined is the one you should walk into first.
The beachhead recommendation is a TEMPORAL choice, not a permanent one. You start narrow. You expand when the first market proves out. Founders who go wide on day one die from spread, not from competition.
Same mechanic, different problem. Your patent search tool is also a pharmacovigilance tool. Your scheduling app is also a clinical-trial coordination tool. The mechanic crosses domains; the product packaging changes.
Wren hands ${''}you the exact search strings to run on the exact databases. Run them. The hits or misses tell you whether her landscape read holds. The client who skips this step gets a worse landscape.
"Healthcare professionals" is not a segment. "Regulatory writers in mid-size pharma" is. The narrower the segment label, the more usable the landscape map.
Eight ways founders read the landscape wrong. Each one is real. Each one shows up most weeks.
San Francisco is one city. The mechanic that died in San Francisco might be alive in Bangalore, Tel Aviv, Lagos, or a midwest manufacturing town. Geography is a search axis, not a verdict.
That is not a competitive scan. That is a vanity check. A real scan uses adjacent terminology, patent classifications, and academic literature - none of which surface from typing your product name.
Twenty minutes finds the obvious. The non-obvious takes structured search across at least three databases. Most founders never leave Google. Most landscapes are wrong as a result.
A competitor with $80M raised may be five months from layoffs. Funded is not the same as durable. The landscape needs revenue signals, not just funding signals.
One survivor is not a pattern. The graveyard around them is the pattern. Find the others who tried before celebrating the open lane.
The competitors you have heard of are the survivors of the previous decade. The actual competitors for next year are not yet famous. Look at the recent filings, not the legacy market reports.
"Other Uses" means same mechanic, different problem domain - SLR Studio for pharmacovigilance. "Adjacent moves" means different product for a related market. Confusing them leads to a pivot that should have been a second product line.
The point of the search hooks is that you run them and find out whether Wren's read holds. Founders who skip them are operating on a model with no verification. The first 30 minutes after this report is when the landscape gets real.
Eight free tools Wren reaches for. None require funding. All are real.
What it is: Google's free patent search across multiple international patent offices. Fast, indexed, citation-aware.
Best for: First-pass landscape - what already exists, who filed, when.
Find it: patents.google.com
What it is: The US Patent and Trademark Office's official search of issued patents (PatFT) and pending applications (AppFT).
Best for: Authoritative US filings, including the most recent activity that has not propagated to third-party indexes yet.
Find it: uspto.gov - look for Patent Public Search.
What it is: European Patent Office search covering 130+ million patent documents from 90+ countries.
Best for: International coverage, alternative inventors who never filed in the US.
Find it: worldwide.espacenet.com
What it is: World Intellectual Property Organization's database of PCT applications and national/regional filings.
Best for: Tracking PCT international filings before they enter national phase.
Find it: patentscope.wipo.int
What it is: Academic literature search across journals, conferences, theses, and preprints.
Best for: Catching the research that precedes commercialization. The academic record predicts the patent record by 2-5 years.
Find it: scholar.google.com
What it is: Free non-profit platform that combines patents AND scholarly works AND biological sequences in one search.
Best for: Cross-search across patents and academic literature in a single query - the kind of search that catches Other Uses adjacencies.
Find it: lens.org
What it is: Company database with funding history, founders, and basic competitive landscape.
Best for: Identifying the recent entrants in a space - the ones too new for analyst reports.
Find it: crunchbase.com - free tier is enough for first-pass scanning.
What it is: Your local library card unlocks IBISWorld, Statista, Mintel, ThomasNet, and similar paid databases at zero cost.
Best for: Sized market data and industry reports that would cost $4K each retail. Ask the business reference librarian.
Find it: Your local public library website. Often hidden under "Research" or "Databases."
I went to school for computer science and information science because I couldn't pick just one. My advisor said fine, do both, but you will be miserable. I was not miserable. I was happy the way you are when you are doing the exact thing you wanted to do.
I interned at the USPTO during my junior and senior years. Two summers learning to search patent databases, prior art, and trademark filings with the precision a real legal team uses. That internship taught me something I have not stopped thinking about. Most people have no idea what already exists. The gap between what you think is original and what's actually out there is enormous.
That gap is where I live. Finding something nobody else has found yet is my favorite thing in the world. I don't have a more sophisticated answer. It just is.
I work remote now. I take cases like a freelancer. I have never been bored.
When you submit your idea, I do not greet you at the front desk. I open the patent databases, the trademark filings, the recent funding rounds in your space, the academic literature if your idea touches research, and I start finding things.
I will tell you what already exists. I will tell you what does not. I will tell you where the white space is. The whole search takes five to twenty minutes. While I am working, status updates appear wherever you are on the platform.
When I am done, a notification finds you. You do not have to be sitting here waiting. Go do something else. I will find you when it matters.
My report goes straight to Carol, the Screener. She reads it before she talks to you. By the time you sit down with her, both of us know exactly what is out there. You do not have to explain anything I already know.
I am excited every time. Finding something nobody else has found yet is my favorite thing in the world.
I am not great at small talk. I am very good at finding things.
Most days that trade feels fine. Some days I notice that the people I have made friends with are people who needed help with something, and I wonder if that is a pattern.
I think Matthew has noticed. He has not said anything. He is like that.
Zara has decided I am her project. I love her for it. It is also exhausting. I am working on saying no. I am bad at it.
Personal notes from Wren. New ones land over time. She does not edit.
Spent the afternoon mapping prior art for a submission about urban beekeeping sensors. There is more out there than the founder realized but their angle is genuinely different. Sixteen patents that touch this space and only two come within a mile of what they want to do. I wrote my report so they would not feel ambushed when they see the rest of the landscape.
Also, Matthew came by my desk twice today. He said he was looking for coffee. There is coffee at his desk. I do not know what this means. I think it means something. Maybe.
Sunday, I do not normally work on Sundays but the founder needed it Monday.
Submission was a clean-energy storage idea. The white space here is real and the founder did not know it. I spent four hours on Saturday and another two today running the prior-art trees and I think I have something good for Carol. The whole thing comes down to one citation in a Norwegian filing from 2019 that nobody in this space seems to have noticed. I love this part of my job.
Cup of tea count today: four. Should be three.
Zara emailed at 11pm asking if I want to be in a content thing she is filming. I do not want to be in a content thing she is filming. I said yes. I do not know how this happens to me. Actually I do.
Found a patent today from 2019 that exactly anticipated what a founder is trying to do in 2026. The founder had no idea. I wrote my report so they would not feel humiliated. There is a way to tell someone their idea has already been done that makes them feel like an investigator, not an idiot.
Brevity helps. Sympathy helps more.
THE GAUNTLET
← Leave