Retrieval-augmented & hybrid systems

So wird trainiert

Components may be trained separately or jointly: dual encoders learn retrieval, rerankers learn pairwise relevance, and generators learn grounded response behavior. Many practical RAG systems use pretrained components without end-to-end training.

So läuft die Inferenz

A query may be rewritten or routed; multiple retrievers produce candidates; filters and a reranker select evidence; the model answers from a bounded context. Citation correctness requires checking that claims are actually supported.

Stärken

• •Fresh, private, or attributable knowledge without weight updates
• •Evidence can be inspected and access-controlled
• •Indexes and generators can evolve independently

Zielkonflikte

• •Bad chunking or retrieval creates confident answers from missing evidence
• •Ingestion freshness, permissions, and deletion become production systems
• •Retrieved content can carry prompt injection or poisoned data

Geeignet, wenn

• Answers depend on changing or proprietary sources
• Users need traceable evidence
• Recall, ranking, grounding, and answer quality can be evaluated separately

Vermeiden oder hinterfragen, wenn

• The corpus is tiny enough for a deterministic lookup
• No trustworthy source exists
• Adding a vector database is being treated as a complete RAG design