Case Studies

Deep dives

Long-form breakdowns of complex system design problems, tradeoffs made, and lessons learned.

Voice / AI Ops 2026-06-02

Voice-Native OS

Speak intent into a microphone. Watch a multi-agent pipeline create a ticket, dispatch it to an isolated executor, verify the output, and merge to main — without touching a keyboard. LiveKit Agents self-hosted + Claude Haiku function-calling + existing tickets.db FSM. ~1.5s end-to-end latency.

Voice dispatch LiveKit FSM Self-hosted Claude Haiku
Read case study →
Architecture 2026-05-07

Master OS

A custom AI operating system spanning 7 business domains and 9 git repos. Phase 6 hybrid orchestration model, hooks-deterministic trust boundaries, ticket-executor FSM, pre-merge-gate, and daily wiki distillation. 241 tests passing, A 19/19 health grade in production.

Orchestration Multi-repo Claude Code Hooks FSM
Read case study →
System Design 2026-05-13

Agentic Orchestration at OS Scale

The architecture decisions behind a multi-agent AI operating system running 7 businesses in parallel: hooks over prompts for invariants, hybrid orchestration forced by runtime constraints, FSM-gated merge pipeline, and distillation as a compounding knowledge mechanism.

Multi-agent Hooks Hybrid orchestration Distillation
Read case study →
Operations / Strategy 2026-05-13

Financial Decisions Framework

A structured framework for capital allocation across a 7-domain portfolio with radically different risk profiles, timelines, and return types. Domain classification, decision type taxonomy, and priority order for cross-domain resource allocation.

Portfolio strategy Capital allocation Operations Frameworks
Read case study →