CAESAR
An agentic workflow evaluation and reliability platform combining generated tests, trace analysis and explicit safety gates.
Portfolio
Selected projects and advisory engagements. Some are open-source, some are client work described at the level of detail that's appropriate.
Selected products, systems, and advisory work across strategy and engineering.
Featured
Evaluation, orchestration and infrastructure for agentic systems.
Additional Work
MealMind AI
An agentic kitchen assistant and macro intelligence system that transforms fridge ingredients into structured recipes and real-time nutritional tracking.
Granite
A local-first thinking system - markdown notes, wikilink navigation, and a calm editorial UI that evolves from linked notes into a durable knowledge graph.
SpecStudio
Native AI IDE that turns intent into specs, plans, and production code through a controlled Intent -> Spec -> Plan -> Code workflow.
TruthTeller AI
Multi-model deliberation platform that routes hard questions through an AI council, then exposes how the final answer was formed.
Soulprint
Consumer AI product that turns playlist exports into psychological profiles and shareable story cards without building a tracking business.
Enterprise AI Governance Framework
Advisory and framework design for enterprise AI controls, decision rights, and assurance models in regulated environments.
Reference Architectures & Patterns
Reusable architecture patterns for agent orchestration, tool safety boundaries, and production monitoring in enterprise AI.