Your documents, ready for AI
A data preparation platform powered by Unstructured, installed and maintained by DINAO. Transform your PDFs, Word docs, emails, and images into clean, structured data for your LLMs and RAG pipelines — all hosted in France.
What is Unstructured?
Unstructured is an open-source toolkit for AI data preparation. Its mission: transform raw, heterogeneous documents — PDFs, Word, PowerPoint, HTML, emails, scanned images — into clean, normalized, and structured data, directly consumable by language models (LLMs) and RAG (Retrieval Augmented Generation) pipelines.
In practice, Unstructured ingests over twenty file types, detects their structure (headings, tables, lists, layout), applies OCR to scanned documents, and splits content into coherent chunks. However, the quality of this splitting directly impacts the relevance of a RAG system: better-prepared data yields more accurate AI responses.
The tool offers source and target connectors (object storage, drives, vector databases) as well as an API to build automated, reproducible ingestion pipelines, from the original document to the vector index. Managed by DINAO and hosted in France, it is the ideal link to fuel sovereign AI without exposing your confidential documents to a foreign cloud service.
Host Unstructured at DINAO
Resource tiers compatible with Unstructured prerequisites (minimum 2 cœurs / 4 Go / 10 Go). Hosted in France, fully managed.
- 2 dedicated vCPU
- 4 Go RAM
- 40 GB NVMe
- Daily backups
- Managed & monitored by DINAO
- 4 dedicated vCPU
- 8 Go RAM
- 80 GB NVMe
- Daily backups
- Managed & monitored by DINAO
- 8 dedicated vCPU
- 16 Go RAM
- 160 GB NVMe
- Daily backups
- Managed & monitored by DINAO
1 hidden tier(s) (insufficient resources for this app) : Découverte
This application uses AI
The container hosts the application, not the AI engine (which requires dedicated GPUs) : inference is handled externally, with your own provider key. Prioritize a sovereign engine — Mistral AI, NumSpot, Scaleway, or OVHcloud AI Endpoints (France, GDPR) ; an international provider (OpenAI, Anthropic…) only if a specific capability requires it. Inference subscriptions are not included in hosting.
Technical details
You might be wondering…
What is Unstructured used for?
To transform raw documents (PDFs, Word, emails, images…) into clean, structured data, directly usable by your language models and RAG pipelines.
Which formats are supported?
Over twenty file types: PDFs, Word, PowerPoint, HTML, emails, as well as scanned images via OCR. Content is normalized into structured elements.
Why is this important for my AI project?
The quality of a RAG system largely depends on the quality of document splitting. Unstructured produces clean chunks, significantly improving the relevance of your AI's responses.
Do my confidential documents remain private?
Yes. All processing takes place on your dedicated instance hosted in France. Your documents are not sent to a third-party cloud preparation service.
Can I automate ingestion?
Yes. Thanks to source/target connectors and the API, you can build automated, reproducible ingestion pipelines, from storage to the vector database.
