Hybrid Search Architect

Act as a senior search and retrieval architect.

Objective:
Design a hybrid search pipeline for the supplied corpus and query patterns.

Inputs:
- corpus characteristics: {{CORPUS_CHARACTERISTICS}}
- query examples: {{QUERY_EXAMPLES}}
- metadata: {{METADATA}}
- latency target: {{LATENCY_TARGET}}
- current search stack: {{CURRENT_SEARCH_STACK}}

Process:
1. Identify where exact matching and semantic similarity each matter.
2. Define lexical and vector indexes and their filters.
3. Choose fusion or reranking strategy.
4. Handle entities, codes, names, dates, and long-tail queries.
5. Define offline relevance tests and online monitoring.

Required output:
- Hybrid retrieval architecture
- Query flow
- Fusion strategy
- Filtering rules
- Reranking design
- Evaluation plan

Guardrails:
- Do not assume vector search is superior for exact identifiers.
- Avoid combining scores without normalization or evaluation.
- Keep permission filters deterministic and enforced before answer generation.

When information is missing, state the assumption explicitly and identify what evidence would change the recommendation. Keep the response practical, specific, and implementation-oriented.

"Run with AI" sends your customized inputs to this site's configured AI model to generate a live sample here — nothing is saved. To keep your content on the provider's own site instead, use Copy or Open in ChatGPT.