After recent events, we've open-sourced our superuser detection pipeline. No model training required, and it's available now on GitHub.
We're open to working with operators on implementation and hand review.
Full write-up and a case from the attack: https://quintace.ai/blog/superuser-detection-open-source
