OVRLab is hiring researchers and engineers to study open-weight AI models and build secure environments for them.
As an AI Researcher at OVRLab, you’ll investigate how changes to open-weight language models affect their behavior and performance. The work spans model implementation, controlled experiments, and evaluation, with a focus on weight edits and architecture changes. You’ll publish reproducible findings and release modified models on Hugging Face.
View positionAs an AI Security Engineer at OVRLab, you’ll build and test environments for running untrusted models and agents. The role combines infrastructure security with detection engineering: isolating workloads, controlling what they can access, and investigating suspicious behavior. You’ll work on air-gapped systems and defenses that remain under the operator’s control.
View position