Reducto on parsing the Epstein files into structured data for JMail
AI Engineer · youtube · 2026-10-11
Reducto's Palak Agarwal and Omar Alhait walk through turning heavily redacted PDFs and handwritten flight logs into structured data powering JMail and JFlights. Covers reading order/handwriting/layout bottlenecks, OCR vs PDF metadata, safe redaction, generating and refining JSON schemas (nested fields, citations, figures), deploying extraction pipelines via API, and a sample-then-benchmark evaluation approach before scaling.
More from coding & agent
- Evals are replays: store tasks in R2, sandbox in Docker, and 90% of the work is measuring — danshipper · 2026-10-11
- Autoresearch agents get stuck in 'idea basins' — fork-and-flush offers a fix — menhguin · 2026-10-11
- JetBrains Air hands-on: running two parallel coding agents inside WebStorm — mhdfaran · 2026-10-11
- Vibe coding an Ori-style platformer with a giant boss chase using Opus 5.5 — chongdashu · 2026-10-11
- Dev builds tiny logger to find what eats OpenAI budget — token counts pointed to the wrong feature — Atm1n9 · 2026-10-11
- Clanker Cloud Adds Email Integration So Agents Can Handle Your Inbox — tekbog · 2026-10-11