RAG-ready documents. No pipeline. No compromise.

EU-hosted. No ML engineer. Upload → structured chunks → done.

Convert your first document free Sign in

Input

Quarterly Report Q3 2024
Finance Department

Total revenue reached €4.2M,
a 12% increase year-over-year.

Output

---
os_component: finance-report
context_tier: summary
domain: finance
token_count: 94
---

# Q3 2024 Revenue Summary

Total revenue reached **€4.2M** (+12% YoY).

Built for European teams

SaaS companies, consultancies, law firms, agencies — any European team that needs RAG-ready output from their documents without an ML engineer, compliance risk, or pipeline complexity.

Your data stays in the EU

Supabase Frankfurt · Mistral EU · Vercel EU. CLOUD Act does not apply.

An appliance, not a pipeline

No model selection. No connector setup. Upload and get chunks back.

Schema-valid output, no ETL needed

Every chunk validated before delivery. Drop directly into LangChain or LlamaIndex.

How it works

Three steps. No pipeline.

Upload

Drop a PDF or DOCX via the web UI or REST API.

distil processes

Mistral AI extracts structure and generates semantic chunks with YAML frontmatter — entirely within the EU.

Download

Receive a validated ZIP of Markdown files. Ready for any vector store.

Schema

Every chunk ships with structured metadata

distil generates YAML frontmatter on every Markdown chunk. Each field is purpose-built for RAG — no post-processing, no custom parsing.

•os_component — Semantic section label derived from document structure

•context_tier — summary | detail | reference — for LLM context budgeting

•domain — Topic classification: finance, legal, hr, technical, …

•token_count — Approximate token count — no re-tokenising needed

•source_page — Source page — full traceability back to the original

•chunk_index — Position in sequence — reconstruct reading order

EXAMPLE CHUNK — contract-review.pdf

---
os_component:  liability-clause  ← section label
context_tier:  detail            ← summary | detail | reference
domain:        legal             ← filtered retrieval
token_count:   187               ← context window budgeting
source_page:   12                ← traceable to original
chunk_index:   8                 ← reading order
---

## 9.2 Limitation of Liability

Neither party shall be liable for indirect,
incidental, or consequential damages arising
out of or related to this Agreement...

Pricing

Flat monthly pricing. No per-page surprises.

Free

€0

50 pages / month

REST API access
ZIP download
YAML frontmatter
Email support

Start free

Built for European compliance from day one.

Frankfurt, DEAll data processed within EU jurisdiction

No training dataYour documents are never used to train or fine-tune any model.

CLOUD Act: N/ANo US entities in the data path

Start converting today

Free tier. No credit card required.

Convert your first document free