Independent technology lab
Independent · We are the enabling partner21M Labs
An independent lab in Denver that opens more files than it finishes, and writes down in advance what would make it stop.
Every venture practice has the same hole in it, which is where the companies come from.
Building a testable version of an idea now takes days. Deciding which ones to stop has not got cheaper at all, and that asymmetry is the whole game. A studio that cannot kill things accumulates a portfolio of survivors that were never tested, and an advisory practice with no origination waits for whatever walks through the door.
The usual answer is to run a lot of experiments and hope. That produces volume, not evidence. What is missing is a gate: something written down before anyone is attached to the answer, that says what would make this file stop, and a date on which somebody actually checks.
The lab opens files. Retention decides which ones become companies.
Every file the lab opens carries a written assumption, a kill condition set on day one by the person who most wants the file to live, and a date to check it against. Most assumptions do not survive the check, and the closing note gets written the same way whether the file earned money or nothing at all. The procedure is four moves: notice what is expensive and tolerated, name the assumption, run it small enough to be wrong quickly, and close the file properly.
That is where Ezra Bridge comes in, and only there. A file that shows real retention, meaning people came back to it without being asked twice, gets funding and venture building through us. The rest stay in the lab or get closed with the reasons published. We are buying evidence, not enthusiasm, and we are buying it from a lab that has already spent its own time disproving most of it.
The lab keeps autonomy over what it opens and which verticals it works in. We do not get a vote on that. The arrangement covers funding and venture building, and it stops there, which is what keeps the origination honest.
Three files, and one of them is live.
BuildIQ was file 001 and is now a venture on this page. Pullup.gg is live: skill-based wagering across thirteen titles where the stake sits in escrow and settles itself, so the operator is never in the money path and cannot stall a payout it is holding. brine is a public randomness beacon that pulls its entropy off the sea state reported by forty-six ocean buoys, published under MIT, on a thirty-minute cadence, with a method anyone can recompute and argue with.
The three have nothing in common except the procedure that produced them, and that is the point being made. The lab is the repeatable part. For Ezra Bridge it is a pipeline of companies that have already been tested to destruction by the people who built them, which is a materially better starting position than an inbound deck.