public_documentation-library_basics

Evie’s Library is a tenant-specific knowledge management system that centralises organisational information into structured **Catalogs** and semantically indexed **Documents** for AI-driven retrieval. It enables context-aware AI interactions by processing content for meaning rather than keywords, supporting multilingual queries and scalable knowledge reuse across workflows.

Evie’s Library is a tenant-specific knowledge management system that centralises organisational information into structured Catalogs and semantically indexed Documents for AI-driven retrieval. It enables context-aware AI interactions by processing content for meaning rather than keywords, supporting multilingual queries and scalable knowledge reuse across workflows.

Core Components

The system organises knowledge into Catalogs (customisable sections like "Client Portfolios" or "Product Lines") and Documents (semantically indexed files such as contracts, brochures, or reports). Processors handle document ingestion (e.g., an automagic HTML processor or customisable alternatives), while Retrievers define how AI Specialists access Catalogs—typically via semantic search. This structure ensures AI responses are grounded in precise, up-to-date organisational context.

Semantic Indexing & Multilingual Support

Documents undergo semantic indexing, splitting content into meaningful chunks for accurate retrieval—even when queries use different terminology. The Library natively supports multilingual operations: documents in any language can coexist in the same Catalog, and queries in one language can retrieve results from another without requiring manual translations.

Document Management Workflows

Documents are added via URLs (enabling automatic refreshes and version tracking) or file uploads. URL-based documents sync with source changes, maintaining currency without manual updates. Processing occurs in the background, with the latest version available once indexing completes. For manual uploads, users should periodically review for updates.

Integration & Scalability

The Library scales with organisational growth, accommodating new Catalogs and Documents without disrupting existing workflows. API access allows programmatic maintenance and integration with external systems. Best practices include grouping related documents, using descriptive metadata, and preferring URL-based sources for automatic updates.

Practical Applications

In sales scenarios, Catalogs might organise Client Profiles, Product Knowledge, and Sales Playbooks, while AI Specialists use indexed Documents to answer queries or support real-time interactions. This reduces manual search time and ensures consistent, personalised responses—enhancing efficiency and client engagement.

The system’s flexibility extends to Dossier Catalogs (with tagging support) and Standard Catalogs, with metadata and naming conventions further optimising discoverability.