Datacurve
About
Builds training and evaluation data, including reinforcement-learning environments and agent trajectories, for frontier model development.
What They Offer
Commercial or operational services this company provides.
Agent evaluations
SFT data
Custom RL environments
Training datasets
Agent trajectories
Products & Public Artifacts
Wide Research
Tool · Open source
Tool for spawning thousands of sub-agents in parallel.
Visit artifact →Pier
Tool · Open source
Harbor fork with stronger support for CLI coding agents in offline environments.
Visit artifact →DeepSWE
Benchmark · Open source
Long-horizon software-engineering benchmark: 113 original tasks across 91 repositories in 5 languages, designed to resist training-data contamination.
Visit artifact →Technical Capabilities
Long-Horizon TasksCode / Terminal Execution
Only publicly documented capabilities are shown. Absence does not imply the capability is unavailable.
Focus Areas
CodingRL EnvironmentsTraining DataEvaluations
Domains
Software EngineeringCybersecurity