Research forreasoning agents.
We’re building the data and environments for AI to learn the judgment behind complex professional work.
Building the foundations of expert intelligence.
Professional expertise takes years to develop. We’re building the training foundations to bring that judgment into AI, across disciplines and beyond isolated tasks.
Task datasets
Context-rich examples of expert work: source material, decisions, constraints, and reference outputs.
Execution environments
Realistic tools and application state for learning complete workflows, including the exceptions and recoveries.
Evaluation criteria
Feedback grounded in source evidence, domain standards, and the quality of the final work.
Intelligence across disciplines.
We’re pursuing agents that can reason across the demands of finance, law, security, and professional practice.
Financial reasoning
Financial models, reconciliations, and analysis grounded in source data.
Legal interpretation
Contract analysis, obligations, and reasoning over referenced clauses.
Security & compliance
Control mapping, security reviews, and SOC 2 evidence.
Privacy & health
Data flows, privacy requirements, and HIPAA evidence.
Workforce operations
Employment policies, people processes, and operational decisions.
Tax & public funding
Tax positions, grant eligibility, and supporting documentation.
Help shape what agents learn next.
Join the beta waitlist for research updates and invitations to work with Meter.