A curated collection of technical deep-dives, research findings, and engineering insights — built in the open.
Deep technical investigations — PDF extraction pipelines, OCR engines, accuracy benchmarks and engineering trade-offs.
Read research →Long-form drafts, refinement iterations and thinking-out-loud posts before they hit the feed.
Browse posts →The full index — everything organized by topic, searchable and always growing as new things get documented.
Open index →