Self Checks
1. Is this request related to a challenge you're experiencing? Tell me about your story.
In the current platform, the RAG knowledge base only provides three text chunking options. We suggest adding mainstream advanced strategies, especially Semantic Chunking, which uses embeddings to measure similarity between adjacent sentences/paragraphs and only splits when the topic actually changes. This keeps chunks more coherent and reduces context fragmentation caused by fixed or rule-only splitting.
We also recommend adding QA-based Chunking / QA Augmentation: during indexing, let an LLM read document sections and generate 3–5 hypothetical questions per section, then store these questions as retrieval anchors. At query time, user questions are matched against these generated QAs, which can significantly improve retrieval precision and intent matching, especially for complex or indirect queries.
2. Additional context or comments
As shown in the screenshot:
3. Can you help us with this feature?
Self Checks
1. Is this request related to a challenge you're experiencing? Tell me about your story.
In the current platform, the RAG knowledge base only provides three text chunking options. We suggest adding mainstream advanced strategies, especially Semantic Chunking, which uses embeddings to measure similarity between adjacent sentences/paragraphs and only splits when the topic actually changes. This keeps chunks more coherent and reduces context fragmentation caused by fixed or rule-only splitting.
We also recommend adding QA-based Chunking / QA Augmentation: during indexing, let an LLM read document sections and generate 3–5 hypothetical questions per section, then store these questions as retrieval anchors. At query time, user questions are matched against these generated QAs, which can significantly improve retrieval precision and intent matching, especially for complex or indirect queries.
2. Additional context or comments
As shown in the screenshot:
3. Can you help us with this feature?