Blog
Engineering deep-dives, release notes, and the decisions behind dOCR — written by the people who build it.
You can now define your own field schemas and have dOCR return exactly those typed fields — nothing more, nothing less. What custom document types are, how to define one, and why they beat generic extraction.
Priya Raghavan
Co-founder & CTO ·
Read post
Why dOCR returns typed, validated JSON instead of a wall of OCR text — and how vision plus an LLM guided by a document-type schema gets you fields you can trust.
Mara Chen
A 12-page invoice is not 12 documents — it is one document with fields scattered across pages and line items that overflow. Here is how dOCR reads the whole thing and returns one typed JSON object.
Mara Chen
Prototype an extraction by hand, lock down the document type that gets it right, then wire it into your backend. The full path from a single upload to extraction at scale.
Priya Raghavan