Work
Published work is produced with AI assistance — research, argument testing, data analysis, drafting, and critique. Each piece is organized so it can be revisited, challenged, and updated.
Contexts
Topics
← All topicsArtificial Intelligence
Life Calibration Engine
A mobile-first PWA and agent-accessible MCP for life assessment, goal prioritization,
OpenCode Go Chat
A browser-based chat interface for OpenCode Go with a deployed proxy for chat, search, fetch, and browsing tools.
LLM Benchmark Platform
A deployed benchmark platform for running role-based coding-agent evaluations across OpenCode Go models and tracking results over time.
Data Analyst AI Agent
AI-powered data analytics dashboard with natural language chat and real-time visualizations
First Supabase-Published Artifact
Proving the canonical path: publish request through the Publishing API into Postgres, rendered without any deployment.
Agent Native Applications
A first-principles definition of agent native applications: software products reorganized around delegated intent, governed AI agents, and supervisory interfaces, with email as the central example.
State of the AI Industry
A living analysis of the AI industry across six layers — from silicon to business domains — written from the vantage point of a Data Scientist who helps organizations make better decisions.
The Economics of LLMs: Training, Inference, and Who Pays
Training a frontier model costs nine figures. Serving it costs fractions of a cent per token. The gap between those two numbers defines who can compete in AI — and who gets left out.