Projects

CLI pipeline that tailors resumes to job descriptions using an LLM for content selection, Typst for PDF compilation, and a metrics-driven fit loop that automatically adjusts content until the output meets page and spacing constraints.

Current State: Working local prototype for recording sessions, dealt games, scoring events, substitutions, history, and sample scenarios. It also has a first shared-scorebook flow backed by Cloudflare Worker + D1.

An agentic RAG pipeline that indexes research papers, answers natural-language queries, and surfaces connections across studies.