What it sells
We find flaws in RL environments and evals and help you fix them.
Business model breakdown
Buyer: who owns the budget?
Teams training models with RL, building environments or running agent evals (labs, eval vendors).
Trigger: what makes them buy now?
Training amplifies flaws in graders; need to separate real capability gains from exploitation.
Workflow: what inputs become what output?
Run an agent swarm (EnvCheck) over RL environment or eval tasks, peer-review findings, report reward hacks and grading bugs to fix; found 110 issues across Terminal-Bench 4.0, DeepSWE 1.1, BFCL v4.
Economics: what it saves and how it charges
Tooling and audit service; blog claims audit costs roughly $3.3-$9.4 per task, similar to running the benchmark task.
Distribution: how it reaches customers
UnverifiedContact-us led, founder outreach and public research blog.
Funding history and coverage
| Round date | Stage | Amount | Investors | Source |
|---|---|---|---|---|
| — | Y Combinator Summer 2026 | — | Y Combinator | ycombinator.com |
Profiles summarize public reporting and company websites. Business-model notes are MaxBC's reading of that material, not confirmed facts; unknowns are left blank rather than guessed. Updated 2026-10-12.