Five jobs Bohrin does today.
Pick the one that sounds like your week. Each is a short read: what goes wrong, what Bohrin does about it, and the real commands and output to try it yourself.
Coding-agent teams01
Weakened tests, caught on the pull request
Your agent says “fixed, all tests pass”. Bohrin shows you whether it fixed the code or changed the tests.
Read the use case4 min read
RL environment builders02
Graders that can’t be passed without the work
Before you ship an environment, find out whether its grader pays for answers that did none of the work.
Read the use case3 min read
Evaluation teams03
Know if +3.5 points is real
Every score with its margin of error, and the smallest difference your evaluation can actually detect.
Read the use case2 min read
Teams fine-tuning with their own reward04
Check the reward before you pay for the run
A reward that only checks the code runs pays full marks for an empty function. Find that before training does.
Read the use case2 min read
AI labs and post-training teams05
Check the environments you buy before you train on them
Ask for results you can re-run yourself, in a shared vocabulary, with what wasn’t checked written down.
Read the use case3 min read