ArchivIA
What is ArchivIA?
ArchivIA is Normadat’s retrieval-augmented generation (RAG) application for querying an organization’s document archive in natural language. It semantically indexes authorized corporate documentation and returns source-cited answers through a Conversational AI interface, operating as an Enterprise Search layer over existing document repositories.
Key Capabilities
- Semantic indexing and natural-language retrieval: ArchivIA indexes documentary collections so users can locate information by meaning and context. Queries can address files (expedientes), contracts, specifications, reports, resolutions, internal procedures, and other documentary types, including unstructured and handwritten material.
- Source-grounded RAG answers: Responses are generated from the organization’s defined documentation. Each answer can include citations, page numbers, and links to the original sources inside the archive.
- Cross-document analysis: ArchivIA identifies information across multiple documents, surfaces recurring criteria and documented exceptions, generates summaries and patterns, and flags typos and missing items in files.
- Internal guide generation: The product can produce internal guides, criterion summaries, and operational orientation from the indexed corpus for technical and administrative staff.
- Organization and metadata: Documents can be organized and filtered by type. ArchivIA supports metadata query, automatic classification, and bulk loading.
- Access control and traceability: Role-based authentication and authorization apply. ArchivIA respects existing documentary permissions so each user reaches only authorized material. Query logs record what was asked and what was returned, supporting internal audit and usage analytics.
- Repository integration: ArchivIA connects to corporate document managers, SharePoint, email, Google Drive, Active Directory, and Samba. Each implementation is configured by documentary typology.
Audience & Use Cases
- Audience: Public administrations, regulated organizations, and teams in legal, compliance, procurement, human resources, back office, and citizen or customer service functions that work with large documentary holdings.
- Use Case: Locate clauses, consult precedents, review administrative files, identify pending documentation, generate knowledge from historical archives, support resolution of similar cases, onboard new staff from existing procedures, and homogenize criteria across large organizations.
Technical Specifications
- Architecture: Cloud-native deployment with per-client isolation. Language models run in a controlled environment on RAG flows connected to authorized sources.
- APIs: APIs implemented with FastAPI.
- Generation scope: Answers are produced exclusively from previously defined corporate documents. ArchivIA does not query external or public sources. Client data remains in the customer environment and is not used to train third-party models.
- Integrations: Corporate document management systems, SharePoint, email, Google Drive, Active Directory, Samba.
- Implementation sequence (vendor methodology): Documentary typology and content-quality analysis; archive organization and permission definition; semantic indexing and RAG configuration on validated documentation; deployment of the natural-language query interface in a controlled environment; ongoing response adjustment.
- Document types: Files, contracts, resolutions, reports, procedures, specifications, and related administrative and legal documentation.
