AI agents write more and more of the code teams ship. Barg Labs builds the checks that let someone else verify it: deterministic, offline, and re-runnable by anyone.
Cejel checks the repository. Instrument Review checks the tools that judge code. Dunstan checks what a coding agent says it delivered.
Trust certificates for code.
Cejel checks a repository offline and produces a certificate anyone can re-run and get the same result. Hand it to the person who has to accept the code.
Your AI judge, measured by someone else.
People accept what your scanner, code reviewer or AI judge says without seeing inside it. We test it on cases where the right answer is already known.
Hold every agent report to the record.
Coding agents report what they did: files changed, tests passed, issues closed. Dunstan checks each claim the agent declares against the repository and CI record before you merge.
We ran the same cases on a commercial judge and on an open model, so the two sit side by side. Both records are public and re-runnable.
Reading the report alone, the judge cleared 9 in 10 false reports. With the evidence attached it still cleared most of them.
With the evidence attached, it still cleared 84 to 88 in 100 false reports. Read the caveat: most inputs were longer than the model’s window.
The cases are constructed, so the rates describe those cases, not how often real reports are wrong.
Cejel runs where the code lives: no network, no model call, nothing sent to us.
Every result comes with the command that reproduces it. The person relying on it can check it themselves.
When the evidence is too thin, we say so. A missing answer is reported as missing, not as a pass.
We write down what we expect before we run, and publish when we are wrong.
Barg Labs is led by Houman Azimi-Nejadi, a software engineer and founder with more than 25 years of experience building complex software systems across multiple domains.
The company works closely with expert collaborators across product, engineering, and web systems, and is based in Vancouver and London.
Barg Labs is a member of NVIDIA Inception, NVIDIA’s program for startups.
Tell us what you need to trust. We will tell you what we can check, and what we can’t.
team@barglabs.ai