Knowledge Hub & RAG
Turn private documents and structured data into grounded, searchable AI context.
Interactive platform map
Architecture in context
The focused blueprint, its required foundation, and declared recommendations.
Knowledge Hub & RAG
Turn private documents and structured data into grounded, searchable AI context.
What this replaces
General-purpose models cannot reliably answer from private organizational material, and keyword search misses semantic relationships.
What your team gains
A private retrieval layer combining structured PostgreSQL data, pgvector similarity search, full-text search and Open WebUI knowledge workflows.
What is inside the blueprint
How it fits the platform
Documents/data -> extraction/chunking -> EmbeddingGemma/embedding endpoint -> PostgreSQL/pgvector + full-text indexes -> retrieval -> Open WebUI/agents.
PostgreSQL, pgvector, EmbeddingGemma, llama.cpp, Open WebUI
- AICORTEX Core PlatformVector and full-text retrieval live in Core's PostgreSQL data tier.
From prerequisites to operation
- Source documents/data
- Chunking/metadata strategy
- Embedding endpoint
- Storage sizing
- Create knowledge schema and vector/full-text indexes
- Deploy embedding endpoint
- Configure ingestion pipeline
- Connect Open WebUI knowledge backend
- Test retrieval and citations
- Chunk size/overlap
- Embedding model/dimensions
- Metadata taxonomy
- Top-k and score thresholds
- Hybrid weighting/reranking
- Document access scope
- Source freshness
- Re-indexing
- Retrieval quality evaluation
- Embedding model migration
- Database backups
What this unlocks with other layers
Institutional Memory
The AI can retrieve documents, recall prior work and trace relationships and provenance.
Research-to-Knowledge
Discovery becomes a retained, structured research corpus instead of disappearing after one chat.