adityamparikh commented on issue #64:
URL: https://github.com/apache/solr-mcp/issues/64#issuecomment-5650654787

   Design summary carried over from the closed draft #53, with the artifact 
names corrected for Spring AI 1.1.x (`spring-ai-vector-store` and 
`spring-ai-model`, not `spring-ai-core`):
   
   - **Foundation:** a `SolrVectorStore` implementing Spring AI's `VectorStore` 
(extending `AbstractObservationVectorStore`), ported from ai-powered-search, 
with `SolrVectorStoreOptions` for vector field, dimension, content field and 
metadata prefix. Embeddings come from whatever `EmbeddingModel` starter is on 
the classpath; the server must start and all current tools must work without 
one.
   - **Tools:** `semantic-search` (KNN over a `DenseVectorField`), 
`hybrid-search` (BM25 plus KNN merged with reciprocal rank fusion, 
over-fetching each leg), and `index-with-embeddings`. The existing `search` 
tool is unchanged.
   - **Solr side:** `DenseVectorField` with HNSW, KNN queries sent by POST, 
filter expressions converted to `fq`.
   - **Phasing:** foundation and options, then the vector store, then semantic 
search plus RRF, then embedding-aware indexing, then integration tests and docs.
   
   The full document is on the closed PR's branch 
(`docs/semantic-hybrid-search-spec`) for anyone picking this up; it should come 
back under `dev-docs/` together with an implementation.
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to