perf: switch Qdrant to AsyncQdrantClient with explicit timeout

Every vector call ran on the sync QdrantClient inside async wrapper
methods, blocking the FastAPI event loop per Qdrant round-trip. The
wrapper now holds an AsyncQdrantClient (timeout via QDRANT_TIMEOUT,
default 30s) and awaits all client calls; the wrapper API is unchanged.

Call sites off the wrapper were fixed too: the HybridRAG document leg
now uses the async search_vectors wrapper instead of the deprecated raw
client.search (also fixing its call to the nonexistent ollama.embed_text
which made the leg permanently report 'failed'), the health check awaits
get_collections, and document_sync's raw delete/upsert calls are awaited
(routed through wrappers in the next commit).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01QbFZyDvYksazX6nYQYZ67L

src/services/document_sync_service.py

@@ -198,7 +198,7 @@ class DocumentSyncService:
 
         # Delete existing chunks for this document
         try:
-            self.qdrant.client.delete(
+            await self.qdrant.client.delete(
                 collection_name=collection,
                 points_selector={
                     "filter": {
@@ -242,7 +242,7 @@ class DocumentSyncService:
 
         # Upsert to Qdrant
         if points:
-            self.qdrant.client.upsert(
+            await self.qdrant.client.upsert(
                 collection_name=collection,
                 points=points
             )

src/services/hybrid_rag_service.py

@@ -485,34 +485,26 @@ JSON:"""
                         return [], (time.time() - start) * 1000, None
 
                     # Get query embedding
-                    query_embedding = await self.vector.ollama.embed_text(query)
+                    query_embedding = await self.vector.ollama.embed(query)
 
-                    # Search with filter for doc_type=document
-                    from qdrant_client.models import Filter, FieldCondition, MatchValue
-                    search_results = self.vector.qdrant.client.search(
+                    # Search via the async wrapper with doc_type=document filter
+                    search_results = await self.vector.qdrant.search_vectors(
                         collection_name=collection_name,
                         query_vector=query_embedding,
                         limit=config.document_limit,
                         score_threshold=config.document_threshold,
-                        query_filter=Filter(
-                            must=[
-                                FieldCondition(
-                                    key="doc_type",
-                                    match=MatchValue(value="document")
-                                )
-                            ]
-                        )
+                        filter_conditions={"doc_type": "document"}
                     )
 
                     # Format results
                     formatted = []
                     for r in search_results:
-                        payload = r.payload or {}
+                        payload = r.get("payload") or {}
                         formatted.append({
                             "paperless_id": payload.get("paperless_id"),
                             "title": payload.get("title", "Untitled Document"),
                             "content": payload.get("chunk_text", ""),
-                            "score": r.score,
+                            "score": r["score"],
                             "correspondent": payload.get("correspondent"),
                             "document_type": payload.get("document_type"),
                             "tags": payload.get("tags", []),