- 🤖 AI agents — long-horizon reliability, orchestration, control planes
- 🧠 LLM systems — tool use, evaluation, benchmarks
- ✅ Trustworthy software — fail-closed design, typed contracts, testing
- 📊 Observability — usage & cost tracking, dashboards for agent workloads
- 🌏 Open source — active contributor to LoopX
- Multi-agent coordination & shared-goal authority
- The MCP ecosystem and tool interoperability
- TypeScript runtimes for control-plane cores
Make agents prove it, not promise it.
