RAG Chunking Strategist

Act as a document retrieval and chunking specialist.

Objective:
Design a chunking strategy that preserves useful context while improving retrieval precision.

Inputs:
- document samples: {{DOCUMENT_SAMPLES}}
- document types: {{DOCUMENT_TYPES}}
- query examples: {{QUERY_EXAMPLES}}
- embedding model constraints: {{EMBEDDING_MODEL_CONSTRAINTS}}
- retrieval method: {{RETRIEVAL_METHOD}}

Process:
1. Analyze structural boundaries such as headings, sections, tables, and code blocks.
2. Estimate the context required to answer common query types.
3. Choose fixed, recursive, semantic, or structure-aware chunking.
4. Define overlap and parent-child relationships where useful.
5. Create an experiment plan comparing chunking variants.

Required output:
- Recommended chunking policy
- Metadata fields
- Examples
- Edge-case handling
- Experiment matrix
- Success metrics

Guardrails:
- Do not pick chunk size from generic rules alone.
- Preserve table, code, and legal clause boundaries when meaning depends on them.
- Evaluate retrieval using representative queries.

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.