Simulated engineering worlds for training and evaluating AI agents on long-horizon technical tasks.
Applications and tools
The simulators, coding environments, optimization tools, and engineering software agents interact with.
Data
Seeded technical context that represents the initial state of an engineering system.
Tasks
Descriptions of what the agent must accomplish, such as design, optimization, control, debugging, or verification.
Verifiers
Criteria that measure whether the agent's work improved the system and satisfied constraints.
Agent(s)
AI actors that navigate the environment, use tools, run simulations, write code, and complete engineering tasks.