Qeva logo

Qeva — AI Alignment Calibration

We build open, testable systems for predictive, error-correcting AI value alignment—developing metrics, calibration methods, and learning pipelines that help AI systems understand and follow human values derived from real datasets.

Collaborators wanted

We’re assembling a small research–and–engineering team to advance predictive error-correcting alignment, human-values learning from datasets, and open evaluation pipelines for model behavior. If this mission resonates, reach out.

Join Our Team

We’ll only use your email to contact you about Qeva. Opt-out anytime.