Sub-Issue 2: Process and Index IPCC Content for Retrieval

[aleenaharoldpeter]

Description

Once IPCC documents are being ingested, their content must be transformed into a retrieval-ready format using the same workflow currently used by CABook where possible.

Tasks

• Evaluate whether the existing CABook chunking strategy can be reused.
• Process IPCC content into chunks.
• Generate embeddings for the chunks.
• Store IPCC chunks in the retrieval index/vector database.
• Verify that indexed content can be searched.

Acceptance Criteria

• IPCC content is chunked successfully.
• Embeddings are generated.
• IPCC content is present in the retrieval index.
• Relevant IPCC passages can be retrieved through search.