← Back to Vertiva Resources
Bottom Line
Federated Search fans every natural-language question out across three complementary retrieval methods — semantic, keyword, and knowledge-graph — and fuses the results into one answer with unified citations.
How It Works: Three Modalities, One Answer
| Modality | Store | Strength It Contributes |
|---|---|---|
| Semantic | Dense vectors (Qdrant) | Meaning and paraphrase — finds relevant content even when wording doesn't match |
| Lexical | Elasticsearch (BM25) | Exact terms, identifiers, and rare tokens |
| Graph | Neo4j (knowledge graph) | Entity relationships and multi-hop reasoning |
Why It Matters
No single retrieval method is good at everything. Semantic search alone misses exact identifiers; keyword search alone misses paraphrased content; graph search alone misses free-text nuance. Fusing all three closes those gaps.
Trust Is Preserved Across All Three Channels
The platform's core access rule — a user can only retrieve content at or below their clearance level — is enforced identically across all three modalities, with unified citations preserving the same provenance regardless of channel.