feat(rag): add OpenVINO NPU embeddings service

docs/swarm-infrastructure.md

@@ -32,7 +32,8 @@ local AI/search/voice services
         +--> SearXNG :18803
         +--> Brave MCP :18802
         +--> llama.cpp :18806
-        +--> Ollama embeddings :18807
+        +--> Ollama embeddings :18807 (legacy/CPU fallback)
+        +--> OpenVINO NPU embeddings :18817
         +--> Kokoro TTS :18805
         +--> Whisper NPU :18816
 ```
@@ -121,7 +122,8 @@ Docker services:
 Host/user services:
 
 - `llama-server.service` — `:18806`, local llama.cpp OpenAI-compatible LLM
-- `ollama.service` — `:18807`, embeddings API
+- `ollama.service` — `:18807`, legacy/CPU embeddings API fallback
+- `openvino-embeddings.service` — `:18817`, OpenVINO NPU embeddings API (`/v1/embeddings`, `/api/embed`, `/api/embeddings`)
 - `docker-health-endpoint.service` — `:18809`, read-only container health for n8n
 - `obsidian-reindex-endpoint.service` — `:18810`, Obsidian/RAG reindex trigger
 - `url-content-extractor.service` — `:18812`, YouTube/PDF/web extraction
@@ -143,7 +145,8 @@ RAG/vector store:
 
 - ChromaDB path: `~/.hermes/data/rag-search/chroma/`
 - Reindex state/progress: `~/.hermes/data/rag-search/obsidian_index_state.json` and `obsidian_reindex_progress.json`
-- Embeddings backend: Ollama on `:18807`, normally `nomic-embed-text`
+- RAG query/reindex embedding backend: still Ollama on `:18807` with `nomic-embed-text` until a deliberate full Chroma rebuild/migration is run.
+- RAG/embedding health probe backend: OpenVINO NPU embeddings service on `:18817`, currently `bge-base-en-v1.5-int8-ov`.
 - Reindex endpoint: `POST :18810/reindex` for incremental updates, `POST :18810/reindex?full=true` for full semantic rebuilds, `GET :18810/semantic-health` to verify vectors plus a search smoke test.
 
 ## Monitoring model