
Lore.
A multi-tenant RAG copilot with embeddable widget, streaming chat, and per-message cost tracking — shipped end-to-end in nine days.
Shipped engagements. Each one fixed-price, fixed-timeline, deployed to production. The case studies tell the rest — the metrics, the trade-offs, the things I'd do differently next time.

A multi-tenant RAG copilot with embeddable widget, streaming chat, and per-message cost tracking — shipped end-to-end in nine days.

A Gmail copilot that classifies, prioritizes, and drafts replies across multiple inboxes — with localized analytics in three languages.

A CLI that scaffolds production-ready Next.js + Supabase + multi-provider AI projects in under two minutes — extracted from the workflow I run on every engagement.

A companion CLI that measures whether your RAG pipeline actually works — retrieval precision, faithfulness, correctness, multi-provider judge LLM.