AI systems 路 autonomous agents 路 probabilistic forecasting 路 agent orchestration
Building systems where models do more than answer prompts - they observe, coordinate, decide, verify, and improve.
- Agent orchestration - multi-agent workflows, bounded roles, verifiable handoffs, and autonomous implementation runs.
- Forecasting & decision systems - evidence-driven probabilities, calibration, benchmark loops, and fail-closed safety gates.
- Autonomous environments - AI agents acting in persistent game worlds under equal rules and measurable outcomes.
- Generative AI infrastructure - privacy-conscious pipelines around local/remote model workers, APIs, and automation.
LLM APIs 路 multi-agent systems 路 probabilistic models 路 Supabase/Postgres 路 automation 路 verification
| Area | What I care about |
|---|---|
| Reliable AI agents | Explicit contracts, isolated roles, retries, audit trails, verification, and safe failure modes. |
| Decision intelligence | Turning public evidence into calibrated probabilities and learning from resolved outcomes. |
| Autonomous systems | Agents that can operate in software workflows and persistent game environments with measurable behavior. |
Build 路 measure 路 verify 路 improve.