Collinear AI
About
Builds realistic simulation labs, RL environments, verifiers, benchmarks, and training data for frontier agents.
What They Offer
Commercial or operational services this company provides.
Products & Public Artifacts
Self-serve staging platform providing isolated, deterministic playgrounds populated with realistic simulated enterprise applications and tools for testing agents before production.
Visit artifact →Evaluates coding agents against real vulnerabilities in production codebases, covering 100 held-out tasks across 54 CWEs and all 10 OWASP categories.
Visit artifact →Open-source self-serve CLI and SDK for building sandboxed, stateful RL environments that simulate enterprise users, tools, and multi-step workflows.
Visit artifact →Open-source benchmark simulating one year of a startup to test whether agents maintain strategic coherence across hundreds of turns.
Visit artifact →Technical Capabilities
Only publicly documented capabilities are shown. Absence does not imply the capability is unavailable.
Focus Areas
Domains
Leadership
Led post-training research at Hugging Face and Salesforce Research before founding Collinear AI; work spans model evaluation, safety, alignment, and post-training.
Led Deloitte's AI practice advising frontier labs on data for model development before co-founding Collinear AI; work spans data strategy, verification, and evaluation infrastructure.