🛠️ Package development: Python package checks, Python standards and workflows, R standards and workflows, Sphinx documentation for R, Package attestation badges, etc.
⚡ LLM infrastructure: Batch inference, Routing-policy evaluation, Capacity planning, Cache-control semantics.
🧪 UI and agent testing: Agent scenario testing, Simulated users and replay, Visual and accessibility checks, Accessibility checks in GitHub Actions, Benchmarking UI judges.
📊 Tabular data: Statistical questions and answers from tables, Learning table knowledge, Disambiguating records, Bloom-filter joins, Data frames with persistent metadata, etc.
📖 Reading, text, and OCR: Academic papers to audio, Audiobooks and reading tools, Combining Wikipedia across languages, OCR, Approximate search and replace, etc.
🌐 Data access: YouTube, YouTube Analytics and Reporting, Weather by zip code, Life tables, etc.
📣 Publishing and discovery: Organization websites, Academic website theme, Classifying starred repositories, Related repositories, Repository statistics, Ghost to Bluesky, etc.
📚 Other: Portable web-app demos, API monetization, Delegated administration prototype, PII-redaction experiments, Agent skills, Mac cleanup, Cross-language puns.
