Getting phone-captured text into your computer: OCR, vision models, agentic pipelines
silenceimpaired · reddit · 2026-09-23
The poster hit the first bottleneck of agentic workflows: reliably importing text captured by phone photos. Auto OCR exists on the phone but the results get lost by the time they reach the computer. The plan: use pi.dev on a watched folder to convert each image to text and stitch it into one final document, possibly with a dual OCR + vision route where an agentic harness votes on the better output. They also want an image model that turns crappy phone shots into clean scanned-looking documents, suspecting image-edit models can partially do this, and are asking the community for tool and model recommendations.
More from coding & agent
- Security audit: autonomous research program XBOW credited with ~12 upstream bug fixes — moyix · 2026-09-23
- Telling agents to use formal verification helps them write better code — sh_reya · 2026-09-23
- Geoffrey Huntley: AI is a time compression device for exploring idea space — kieranklaassen · 2026-09-23
- Team uses Typesafe's Jev to prompt follow-up questions during ticket creation — TheMoonMidas · 2026-09-23
- The most common evals mistake: skipping error discovery and measuring the wrong thing — FinanceYF5 · 2026-09-23
- AI products are easy to change and hard to predict: evals turn 'good' into repeatable tests — FinanceYF5 · 2026-09-23