Hub documentaire

The Document Hub is the central repository for all files in a project. Every document you upload is automatically processed — text is extracted via OCR if needed, chunked into semantically meaningful segments, and embedded for fast AI retrieval.

Once a document is processed, it is immediately available for querying in the AI Assistant, template filling, document generation, and use in analysis tools like Change Impact.

You can upload documents to a project in several ways: by dragging files directly onto the upload area, by clicking the "Upload" button and selecting files from your computer, or by syncing from a connected cloud storage provider (SharePoint, Google Drive, Procore).

1. 1

   Ouvrir le projet

   Naviguez vers Projets dans la barre latérale et ouvrez le projet dans lequel vous voulez uploader.

2. 2

   Uploader vos fichiers

   Faites glisser et déposez des fichiers sur la zone d'upload, ou cliquez sur « Upload » et sélectionnez des fichiers depuis votre ordinateur. Vous pouvez uploader plusieurs fichiers à la fois.

3. 3

   Attendre le traitement

   Alloovium traite chaque document automatiquement. Vous verrez un indicateur de progression à côté de chaque fichier. Le traitement prend généralement 30–120 secondes selon la taille du fichier.

4. 4

   Commencer les requêtes

   Une fois le traitement terminé, le statut du document passe à « Prêt ». Vous pouvez maintenant l'interroger via l'Assistant IA.

Documents are always organised inside a project. A project is a workspace that holds a set of related documents — for example, all documents related to a specific contract or tender. You can create as many projects as you need and invite different team members to each one.

Within a project, you can create folders to further organise your documents. Folders are optional — all documents in a project are available for querying regardless of which folder they are in.

If you have a loose batch of mixed files with no folder structure, use the AI Organizer from the main Projects page. It previews a project-and-folder layout before anything is created, so you can sort a messy intake of specifications, drawings, registers, memos, standards, and instructions in one pass.

1. 1

   Ouvrir Projets et démarrer l'Organisateur IA

   Depuis la page principale Projets, cliquez sur « Organisateur IA ». Ce flux est conçu pour les lots non triés qui ne sont pas encore dans un projet.

2. 2

   Uploader les fichiers en vrac

   Déposez l'ensemble du lot d'un coup. L'organisateur lit les noms de fichiers, les chemins et le contenu extrait pour déduire les projets et types de documents probables.

3. 3

   Examiner la structure proposée

   Alloovium montre les projets suggérés et les dossiers cibles avant l'upload. Les fichiers à faible confiance sont signalés pour vérification manuelle.

Alloovium automatically classifies uploaded documents into categories to help with organisation and to optimise analysis. Common types include contracts, specifications, drawings, reports, and correspondence. You can also assign a custom type to any document.

Understanding how Alloovium processes your documents helps you get the most out of the platform. Each document goes through the following stages:

1. Extraction de texte — Alloovium reads the raw text from each page. For scanned PDFs and images, GPU-accelerated OCR is applied automatically.
2. Analyse de mise en page — The document structure is analysed: headings, tables, figures, and paragraph blocks are identified and tagged.
3. Chunking — The document is split into semantically meaningful segments (typically 200–500 tokens each) that are small enough for precise retrieval.
4. Embedding — Each chunk is converted into a vector embedding using a high-dimensional model trained on technical and legal language.
5. Indexation — Embeddings are stored in a vector index for fast similarity search, enabling sub-second retrieval during AI queries.