Ragie

The Context Engine for agents, assistants and apps

[Ragie @ragieai]

Scanned checks are one of the hardest documents to parse. Handwriting, inconsistent layouts, security patterns, occasional upside down pages.

A customer ran Ragie Parse on a particularly difficult batch last week. This is what it was built for.

Learn more at https://t.co/nFTHxTujum

[Ragie @ragieai]

It's tax day. Vibe code your own tax assistant with Ragie's new skill.
https://t.co/YeeIkTG9RR

npx skills add ragieai/skills

Full tutorial: https://t.co/ctooEtIiWd

[Ragie @ragieai]

New in the changelog: HTML, EPUB, and Excel files all handle edge cases more reliably now, password-protected PDFs are caught immediately with a clear response, and documents paused due to insufficient credits resume automatically once credits are added.

Stay up to date: https://t.co/gVrNnHklYW

[Ragie @ragieai]

Ragie Parse now supports Agentic OCR, available in beta.

Standard parsing misses a lot: forms, stamps, signatures, handwriting, logos, QR codes. Agentic OCR uses vision models to extract these as structured elements via the Elements API.

The highest accuracy option for visually complex documents.

Docs: https://t.co/LXZLIrJ4It
Stay up to date: https://t.co/gVrNnHklYW

[Ragie @ragieai]

Ragie Document Elements API is now available.

GET /documents/{document_id}/elements returns every structured element from a document in reading order — titles, tables, images, code blocks, and more.

Each element includes its type, text, markdown, page location, and bounding box. Filter by element type or index range.

Stay up to date: https://t.co/gVrNnHklYW

[Ragie @ragieai]

Ragie Document Elements API is now available.

GET /documents/{document_id}/elements returns every structured element from a document in reading order — titles, tables, images, code blocks, and more.

Each element includes its type, text, markdown, page location, and bounding box. Filter by element type or index range.

Stay up to date: https://t.co/gVrNnHklYW

[Ragie @ragieai]

Ragie Instructions now support Context Templates.

You can now prepend document metadata to each extraction call, giving the model richer context about what it's processing before it begins.

If you're building agents that extract structured data across multi-document workflows, this is the kind of control that makes results more accurate.

Stay up to date: https://t.co/gVrNnHklYW

[Ragie @ragieai]

Ragie Parse now handles Form & Signature Extraction.

Captures handwritten signatures, detects signed vs. unsigned, classifies form fields by type, and pulls key-value pairs from invoices, contracts, and more.

Better parsed docs = better context for your agents.

Learn more https://t.co/nFTHxTujum
Stay up to date https://t.co/gVrNnHklYW

[Ragie @ragieai]

Introducing Ragie Parse (Early Access) — a new Agentic OCR pipeline that extracts structured content with higher fidelity than traditional OCR. 25+ element types out of the box: tables, forms, signatures, key-value pairs, barcodes, stamps, and more.

Learn more at https://t.co/1o2WMiwjwU

[Ragie @ragieai]

Spreadsheets now extract embedded images and charts alongside table data. More complete extraction, out of the box.
Stay up to date ↓ https://t.co/yOGWhphWoT

[Ragie @ragieai]

PowerPoint support, improved — PPTX and PPT files now benefit from the same high-resolution extraction pipeline as native PDFs.

Stay up to date ↓ https://t.co/yOGWhphWoT

[Ragie @ragieai]

By popular demand — the file size limit for URL-based document uploads is now 100 MB, up from 60 MB.

Stay up to date → https://t.co/yOGWhphWoT

[Ragie @ragieai]

Coming in Ragie 2.0: visual mapping to the exact place in the spreadsheet where an agent found its answer.

Sneak peek:

[Ragie @ragieai]

Someone removed the context from our billboard.

And… they accidentally made our point.

When AI loses context, it hallucinates.

When our billboard got tagged and stripped of context, it did the same, leaving behind a mysterious AI green rectangle and some very confused graffiti by the Caltrain station.

Fortunately, Ragie is built for exactly these moments: when missing context breaks the experience.

Our billboard didn’t survive the night.

Your Ragie pipelines will. 🚀

[Ragie @ragieai]

A lot of RAG demos look great on day one. Production is where things break. Ragie is built for the part after the demo, the part where context has to stay connected, reliable, and relevant.