Ulam
About
Ulam develops research-grade reasoning data, private evaluations, and verifier-backed RL gyms for mathematics, code, algorithms, and science. Its data includes expert-reviewed reasoning traces designed for post-training and model improvement.
What They Offer
Commercial or operational services this company provides.
Agent evaluations
Rubrics / verifiers
Human expert data
Custom RL environments
Training datasets
Products & Public Artifacts
ErdősBench
Benchmark · Commercial
Research benchmark for evaluating advanced reasoning.
Visit artifact →SimoBench
Benchmark · Commercial
Research benchmark for evaluating advanced reasoning capabilities.
Visit artifact →RL Gyms
Environment · Commercial
Verifier-backed reinforcement-learning environments for reasoning tasks.
Visit artifact →Technical Capabilities
Human / Expert FeedbackLong-Horizon TasksProgrammatic VerifiersStateful / Persistent Environments
Only publicly documented capabilities are shown. Absence does not imply the capability is unavailable.
Focus Areas
RL EnvironmentsTraining DataRLHF / Post-trainingEvaluationsHuman Data
Domains
Software EngineeringScience
Leadership
Przemek Chojecki
Founder