diff --git a/docs/configuration.md b/docs/configuration.md
index f6f5d8e0..0214bf50 100644
--- a/docs/configuration.md
+++ b/docs/configuration.md
@@ -329,6 +329,22 @@ OLLAMA_EMBEDDING_MODEL=all-minilm
 - **Switching models requires re-embedding** all documents (may take time for large note collections)
 - **Old collection remains** in Qdrant and can be deleted manually if no longer needed
 
+#### Startup migrations on existing collections
+
+On the first call to `get_qdrant_client()` against an existing collection, the
+server runs two idempotent migrations:
+
+1. **Payload-index creation** — adds `KEYWORD` payload indexes for `doc_id`,
+   `user_id`, and `doc_type`. Required by Qdrant for any `FieldCondition`
+   filter. Cheap; runs even on healthy collections.
+2. **`doc_id` backfill** — rewrites any legacy integer `doc_id` payloads to
+   strings so they match the keyword index. Skipped after a quick sample
+   shows the collection is already clean. On a dirty collection, the full
+   scroll runs once and writes are emitted point-by-point — expect a delay
+   proportional to point count on the first startup after the upgrade.
+
+Both steps emit INFO-level log lines so operators can track progress.
+
 #### Explicit Override
 
 Set `QDRANT_COLLECTION` to use a specific collection name:
diff --git a/nextcloud_mcp_server/api/visualization.py b/nextcloud_mcp_server/api/visualization.py
index 242c7c13..18bcbd4f 100644
--- a/nextcloud_mcp_server/api/visualization.py
+++ b/nextcloud_mcp_server/api/visualization.py
@@ -511,8 +511,8 @@ async def get_chunk_context(request: Request) -> JSONResponse:
                 raise ValueError("end must be greater than start")
         except ValueError as e:
             return JSONResponse({"success": False, "error": str(e)}, status_code=400)
-        # Convert doc_id to int if possible (most IDs are int)
-        doc_id_val: str | int = int(doc_id) if doc_id.isdigit() else doc_id
+        # doc_id is keyword-indexed in Qdrant as str — pass through verbatim
+        # (no int coercion; producers always stringify on write).
 
         # Get Nextcloud host from OAuth context
         oauth_ctx = request.app.state.oauth_context
@@ -537,7 +537,7 @@ async def get_chunk_context(request: Request) -> JSONResponse:
             chunk_context = await get_chunk_with_context(
                 nc_client=nc_client,
                 user_id=user_id,
-                doc_id=doc_id_val,
+                doc_id=doc_id,
                 doc_type=doc_type,
                 chunk_start=start,
                 chunk_end=end,
@@ -570,7 +570,7 @@ async def get_chunk_context(request: Request) -> JSONResponse:
                         must=[
                             get_placeholder_filter(),
                             FieldCondition(
-                                key="doc_id", match=MatchValue(value=doc_id_val)
+                                key="doc_id", match=MatchValue(value=doc_id)
                             ),
                             FieldCondition(
                                 key="user_id", match=MatchValue(value=user_id)
diff --git a/nextcloud_mcp_server/auth/viz_routes.py b/nextcloud_mcp_server/auth/viz_routes.py
index d072c373..01b8a917 100644
--- a/nextcloud_mcp_server/auth/viz_routes.py
+++ b/nextcloud_mcp_server/auth/viz_routes.py
@@ -285,7 +285,11 @@ async def vector_visualization_search(request: Request) -> JSONResponse:
                         vector = point.vector
 
                     if vector is not None and point.payload:
-                        doc_id = point.payload.get("doc_id")
+                        # SearchResult.id is str; coerce payload doc_id to match
+                        # so the tuple lookup below succeeds even on legacy
+                        # int-typed payloads written before normalization.
+                        raw_doc_id = point.payload.get("doc_id")
+                        doc_id = None if raw_doc_id is None else str(raw_doc_id)
                         chunk_start = point.payload.get("chunk_start_offset")
                         chunk_end = point.payload.get("chunk_end_offset")
                         chunk_key = (doc_id, chunk_start, chunk_end)
@@ -555,8 +559,7 @@ async def chunk_context_endpoint(request: Request) -> JSONResponse:
 
         start = int(start_str)
         end = int(end_str)
-        # Convert doc_id to int (all document types use int IDs)
-        doc_id_int = int(doc_id)
+        # doc_id is keyword-indexed in Qdrant as str — pass through verbatim.
 
         user_id = request.user.display_name
         settings = get_settings()
@@ -580,7 +583,7 @@ async def chunk_context_endpoint(request: Request) -> JSONResponse:
             chunk_context = await get_chunk_with_context(
                 nc_client=nc_client,
                 user_id=user_id,
-                doc_id=doc_id_int,
+                doc_id=doc_id,
                 doc_type=doc_type,
                 chunk_start=start,
                 chunk_end=end,
@@ -620,7 +623,7 @@ async def chunk_context_endpoint(request: Request) -> JSONResponse:
                         must=[
                             get_placeholder_filter(),
                             FieldCondition(
-                                key="doc_id", match=MatchValue(value=doc_id_int)
+                                key="doc_id", match=MatchValue(value=doc_id)
                             ),
                             FieldCondition(
                                 key="user_id", match=MatchValue(value=username)
diff --git a/nextcloud_mcp_server/models/semantic.py b/nextcloud_mcp_server/models/semantic.py
index 4612d26d..bb0091dc 100644
--- a/nextcloud_mcp_server/models/semantic.py
+++ b/nextcloud_mcp_server/models/semantic.py
@@ -11,10 +11,12 @@ class SemanticSearchResult(BaseModel):
     id: int = Field(
         description=(
             "Document ID. Numeric for all currently indexed types (notes, files, "
-            "deck cards, news items). The internal SearchResult.id is typed as "
-            "int|str to leave room for future doc types with string identifiers; "
-            "the MCP response narrows to int and a future widening here would be "
-            "a deliberate, breaking-by-design API change."
+            "deck cards, news items). The internal SearchResult.id is stringified "
+            "for Qdrant's keyword-indexed doc_id payload; the MCP response narrows "
+            "back to int via int(r.id). A future doc_type with non-numeric ids "
+            "would surface here as a TypeError at the narrowing boundary, "
+            "forcing a deliberate widening of this field rather than a silent "
+            "API change."
         )
     )
     doc_type: str = Field(
diff --git a/nextcloud_mcp_server/search/algorithms.py b/nextcloud_mcp_server/search/algorithms.py
index 2657bb45..1d14fdf7 100644
--- a/nextcloud_mcp_server/search/algorithms.py
+++ b/nextcloud_mcp_server/search/algorithms.py
@@ -132,9 +132,12 @@ class SearchResult:
     """A single search result with metadata and score.
 
     Attributes:
-        id: Document ID. Numeric for indexed types today (notes, files,
-            deck cards, news items), but typed as ``int | str`` to allow
-            future doc types that use string identifiers (e.g., file paths).
+        id: Document ID — always a string. Producers stringify their native
+            ID before writing to Qdrant so the keyword payload index on
+            ``doc_id`` matches every point regardless of source doc_type.
+            Public response models (e.g. ``SemanticSearchResult``) re-narrow
+            this back to ``int`` at the MCP boundary via ``int(r.id)`` —
+            see ``server/semantic.py`` for the narrowing site.
         doc_type: Document type (note, file, calendar, contact, etc.)
         title: Document title
         excerpt: Content excerpt showing match context
@@ -151,7 +154,7 @@ class SearchResult:
         point_id: Qdrant point ID for batch vector retrieval (None if not from Qdrant)
     """
 
-    id: int | str
+    id: str
     doc_type: str
     title: str
     excerpt: str
diff --git a/nextcloud_mcp_server/search/bm25_hybrid.py b/nextcloud_mcp_server/search/bm25_hybrid.py
index c5848450..c6f8dfda 100644
--- a/nextcloud_mcp_server/search/bm25_hybrid.py
+++ b/nextcloud_mcp_server/search/bm25_hybrid.py
@@ -208,8 +208,9 @@ class BM25HybridSearchAlgorithm(SearchAlgorithm):
             for result in search_response.points:
                 if result.payload is None:
                     continue
-                # doc_id can be int (files) or str (notes/news_items/deck_cards) — see scanner.py
-                doc_id = result.payload["doc_id"]
+                # doc_id is always str post-normalization, but defensively coerce
+                # legacy int payloads on read until the backfill has run everywhere.
+                doc_id = str(result.payload["doc_id"])
                 doc_type = result.payload.get("doc_type", "note")
                 chunk_start = result.payload.get("chunk_start_offset")
                 chunk_end = result.payload.get("chunk_end_offset")
diff --git a/nextcloud_mcp_server/search/context.py b/nextcloud_mcp_server/search/context.py
index cff7d1cc..c249bb73 100644
--- a/nextcloud_mcp_server/search/context.py
+++ b/nextcloud_mcp_server/search/context.py
@@ -20,7 +20,7 @@ logger = logging.getLogger(__name__)
 
 
 async def _get_chunk_from_qdrant(
-    user_id: str, doc_id: int, doc_type: str, chunk_start: int, chunk_end: int
+    user_id: str, doc_id: str, doc_type: str, chunk_start: int, chunk_end: int
 ) -> str | None:
     """Retrieve full chunk text from Qdrant payload.
 
@@ -87,7 +87,7 @@ async def _get_chunk_from_qdrant(
 
 
 async def _get_chunk_by_index_from_qdrant(
-    user_id: str, doc_id: int, doc_type: str, chunk_index: int
+    user_id: str, doc_id: str, doc_type: str, chunk_index: int
 ) -> str | None:
     """Retrieve chunk text by chunk_index from Qdrant payload.
 
@@ -145,7 +145,7 @@ async def _get_chunk_by_index_from_qdrant(
 
 
 async def _get_file_path_from_qdrant(
-    user_id: str, file_id: int, chunk_start: int, chunk_end: int
+    user_id: str, file_id: str, chunk_start: int, chunk_end: int
 ) -> str | None:
     """Resolve file_id to file_path by querying Qdrant payload.
 
@@ -202,7 +202,7 @@ async def _get_file_path_from_qdrant(
 
 
 async def _get_deck_metadata_from_qdrant(
-    user_id: str, card_id: int
+    user_id: str, card_id: str
 ) -> dict[str, int] | None:
     """Retrieve board_id and stack_id for a deck card from Qdrant payload.
 
@@ -288,7 +288,7 @@ class ChunkContext:
 async def get_chunk_with_context(
     nc_client: NextcloudClient,
     user_id: str,
-    doc_id: str | int,
+    doc_id: str,
     doc_type: str,
     chunk_start: int,
     chunk_end: int,
@@ -306,7 +306,7 @@ async def get_chunk_with_context(
     Args:
         nc_client: Authenticated Nextcloud client
         user_id: User ID who owns the document
-        doc_id: Document ID (int for notes/files)
+        doc_id: Document ID (str — keyword-indexed in Qdrant payload)
         doc_type: Type of document ("note", "file", etc.)
         chunk_start: Character offset where chunk starts
         chunk_end: Character offset where chunk ends
@@ -319,17 +319,10 @@ async def get_chunk_with_context(
         ChunkContext with expanded context and markers, or None if document
         cannot be retrieved
     """
-    # Convert doc_id to int for Qdrant query
-    doc_id_int = (
-        int(doc_id)
-        if isinstance(doc_id, str) and doc_id.isdigit()
-        else (doc_id if isinstance(doc_id, int) else None)
-    )
-
     # Try to get chunk from Qdrant first (fast path)
-    if doc_id_int is not None:
+    if doc_id:
         chunk_text = await _get_chunk_from_qdrant(
-            user_id, doc_id_int, doc_type, chunk_start, chunk_end
+            user_id, doc_id, doc_type, chunk_start, chunk_end
         )
         if chunk_text:
             logger.info(
@@ -350,7 +343,7 @@ async def get_chunk_with_context(
             # Fetch previous chunk if not first chunk
             if chunk_index > 0:
                 before_chunk = await _get_chunk_by_index_from_qdrant(
-                    user_id, doc_id_int, doc_type, chunk_index - 1
+                    user_id, doc_id, doc_type, chunk_index - 1
                 )
                 if before_chunk:
                     # Remove overlap: the last chunk_overlap chars of previous chunk
@@ -371,7 +364,7 @@ async def get_chunk_with_context(
             # Fetch next chunk if not last chunk
             if chunk_index < total_chunks - 1:
                 after_chunk = await _get_chunk_by_index_from_qdrant(
-                    user_id, doc_id_int, doc_type, chunk_index + 1
+                    user_id, doc_id, doc_type, chunk_index + 1
                 )
                 if after_chunk:
                     # Remove overlap: the first chunk_overlap chars of next chunk
@@ -422,9 +415,10 @@ async def get_chunk_with_context(
         f"(Qdrant cache miss, possibly legacy data)"
     )
 
-    # For files, retrieve file_path from Qdrant payload
+    # For files, the doc_id is the numeric file ID (as a string) — resolve it
+    # to a WebDAV path so _fetch_document_text can retrieve the binary content.
     resolved_doc_id = doc_id
-    if doc_type == "file" and isinstance(doc_id, int):
+    if doc_type == "file":
         file_path = await _get_file_path_from_qdrant(
             user_id, doc_id, chunk_start, chunk_end
         )
@@ -498,7 +492,7 @@ async def get_chunk_with_context(
 
 
 async def _fetch_document_text(
-    nc_client: NextcloudClient, doc_id: str | int, doc_type: str, user_id: str
+    nc_client: NextcloudClient, doc_id: str, doc_type: str, user_id: str
 ) -> str | None:
     """Fetch full text content of a document.
 
@@ -590,7 +584,7 @@ async def _fetch_document_text(
             # Try to get board_id/stack_id from Qdrant metadata (O(1) lookup)
             # Otherwise fall back to iteration (legacy data)
             card = None
-            deck_metadata = await _get_deck_metadata_from_qdrant(user_id, int(doc_id))
+            deck_metadata = await _get_deck_metadata_from_qdrant(user_id, doc_id)
 
             if deck_metadata:
                 # Fast path: Direct lookup with known board_id/stack_id
diff --git a/nextcloud_mcp_server/search/semantic.py b/nextcloud_mcp_server/search/semantic.py
index c01b0a37..d391b653 100644
--- a/nextcloud_mcp_server/search/semantic.py
+++ b/nextcloud_mcp_server/search/semantic.py
@@ -140,8 +140,9 @@ class SemanticSearchAlgorithm(SearchAlgorithm):
         for result in search_response.points:
             if result.payload is None:
                 continue
-            # doc_id can be int (notes) or str (files - file paths)
-            doc_id = result.payload["doc_id"]
+            # doc_id is always str post-normalization, but defensively coerce
+            # legacy int payloads on read until the backfill has run everywhere.
+            doc_id = str(result.payload["doc_id"])
             doc_type = result.payload.get("doc_type", "note")
             chunk_start = result.payload.get("chunk_start_offset")
             chunk_end = result.payload.get("chunk_end_offset")
diff --git a/nextcloud_mcp_server/search/verification.py b/nextcloud_mcp_server/search/verification.py
index bb3fb678..e5665f3e 100644
--- a/nextcloud_mcp_server/search/verification.py
+++ b/nextcloud_mcp_server/search/verification.py
@@ -565,7 +565,7 @@ async def verify_search_results(
     # complete by the time `verify_search_results` returns.
     if evict_on_missing and inaccessible:
 
-        async def evict(doc_id: int | str, doc_type: str) -> None:
+        async def evict(doc_id: str, doc_type: str) -> None:
             try:
                 await delete_document_points(doc_id, doc_type, user_id)
             except Exception as e:
diff --git a/nextcloud_mcp_server/server/semantic.py b/nextcloud_mcp_server/server/semantic.py
index c1beb208..e8d1a396 100644
--- a/nextcloud_mcp_server/server/semantic.py
+++ b/nextcloud_mcp_server/server/semantic.py
@@ -224,12 +224,11 @@ def configure_semantic_tools(mcp: FastMCP):
             search_results = verified_results[:limit]
 
             # Convert SearchResult objects to SemanticSearchResult for response.
-            # SearchResult.id is typed `int | str` for forward-compat with future
-            # doc_types, but every currently indexed type uses numeric ids and
-            # the MCP response model narrows to `int`. Casting here makes the
-            # narrowing explicit and surfaces any future string-id type as a
-            # loud failure at the boundary instead of silently widening the
-            # public API.
+            # SearchResult.id is `str` (Qdrant keyword-indexed payload), but
+            # every currently indexed type uses numeric ids and the MCP response
+            # model narrows to `int`. Casting here makes the narrowing explicit
+            # and surfaces any future non-numeric-id type as a loud failure at
+            # the boundary instead of silently widening the public API.
             results = []
             for r in search_results:
                 try:
@@ -304,7 +303,10 @@ def configure_semantic_tools(mcp: FastMCP):
                             chunk_context = await get_chunk_with_context(
                                 nc_client=client,
                                 user_id=username,
-                                doc_id=result.id,
+                                # SemanticSearchResult.id is the int-narrowed
+                                # public form; get_chunk_with_context queries
+                                # Qdrant where doc_id is keyword-indexed as str.
+                                doc_id=str(result.id),
                                 doc_type=result.doc_type,
                                 chunk_start=result.chunk_start_offset,
                                 chunk_end=result.chunk_end_offset,
diff --git a/nextcloud_mcp_server/vector/eviction.py b/nextcloud_mcp_server/vector/eviction.py
index 44874e08..8c7555a7 100644
--- a/nextcloud_mcp_server/vector/eviction.py
+++ b/nextcloud_mcp_server/vector/eviction.py
@@ -18,7 +18,7 @@ logger = logging.getLogger(__name__)
 
 
 async def delete_document_points(
-    doc_id: str | int,
+    doc_id: str,
     doc_type: str,
     user_id: str,
 ) -> None:
@@ -29,7 +29,7 @@ async def delete_document_points(
     not present — Qdrant returns successfully with zero points affected.
 
     Args:
-        doc_id: Document ID (int for notes/files/cards/news, str otherwise)
+        doc_id: Document ID (str — keyword-indexed in Qdrant payload)
         doc_type: Document type (note, file, deck_card, news_item)
         user_id: Owner of the points being evicted
 
diff --git a/nextcloud_mcp_server/vector/placeholder.py b/nextcloud_mcp_server/vector/placeholder.py
index fcd12bca..4950c2fc 100644
--- a/nextcloud_mcp_server/vector/placeholder.py
+++ b/nextcloud_mcp_server/vector/placeholder.py
@@ -31,7 +31,7 @@ from nextcloud_mcp_server.vector.qdrant_client import get_qdrant_client
 logger = logging.getLogger(__name__)
 
 
-def _generate_placeholder_id(doc_type: str, doc_id: str | int) -> str:
+def _generate_placeholder_id(doc_type: str, doc_id: str) -> str:
     """Generate deterministic UUID for placeholder point.
 
     Args:
@@ -46,7 +46,7 @@ def _generate_placeholder_id(doc_type: str, doc_id: str | int) -> str:
 
 
 async def write_placeholder_point(
-    doc_id: str | int,
+    doc_id: str,
     doc_type: str,
     user_id: str,
     modified_at: int,
@@ -60,7 +60,7 @@ async def write_placeholder_point(
     processing completes.
 
     Args:
-        doc_id: Document ID (int for notes/files)
+        doc_id: Document ID (always str — see DocumentTask)
         doc_type: Document type (note, file, etc.)
         user_id: User ID who owns the document
         modified_at: Document modification timestamp
@@ -135,7 +135,7 @@ async def write_placeholder_point(
 
 
 async def query_document_metadata(
-    doc_id: str | int,
+    doc_id: str,
     doc_type: str,
     user_id: str,
 ) -> dict | None:
@@ -185,7 +185,7 @@ async def query_document_metadata(
 
 
 async def delete_placeholder_point(
-    doc_id: str | int,
+    doc_id: str,
     doc_type: str,
     user_id: str,
 ) -> None:
@@ -230,7 +230,7 @@ async def delete_placeholder_point(
 
 
 async def update_placeholder_status(
-    doc_id: str | int,
+    doc_id: str,
     doc_type: str,
     user_id: str,
     status: str,
diff --git a/nextcloud_mcp_server/vector/qdrant_client.py b/nextcloud_mcp_server/vector/qdrant_client.py
index b2aa0e65..a192eeec 100644
--- a/nextcloud_mcp_server/vector/qdrant_client.py
+++ b/nextcloud_mcp_server/vector/qdrant_client.py
@@ -3,7 +3,8 @@
 import logging
 
 from qdrant_client import AsyncQdrantClient, models
-from qdrant_client.models import Distance, VectorParams
+from qdrant_client.http.exceptions import UnexpectedResponse
+from qdrant_client.models import Distance, PayloadSchemaType, VectorParams
 
 from nextcloud_mcp_server.config import get_settings
 from nextcloud_mcp_server.embedding import get_embedding_service
@@ -11,10 +12,137 @@ from nextcloud_mcp_server.embedding import get_embedding_service
 logger = logging.getLogger(__name__)
 
 
+# Payload fields filtered by exact-match in scanner/processor/placeholder/eviction.
+# Qdrant requires a payload index for any field used in a FieldCondition; without
+# one, queries fail with HTTP 400 ("Index required but not found"). All three
+# carry string values after producer normalization, so a KEYWORD index is the
+# correct schema (see ADR notes in commit message).
+_KEYWORD_PAYLOAD_FIELDS: tuple[str, ...] = ("doc_id", "user_id", "doc_type")
+
 # Singleton instance
 _qdrant_client: AsyncQdrantClient | None = None
 
 
+async def _ensure_keyword_payload_indexes(
+    client: AsyncQdrantClient, collection_name: str
+) -> None:
+    """Create KEYWORD payload indexes for fields used in exact-match filters.
+
+    Idempotent: tolerates 'already exists' errors so it can run on every
+    startup against existing collections.
+    """
+    for field in _KEYWORD_PAYLOAD_FIELDS:
+        try:
+            await client.create_payload_index(
+                collection_name=collection_name,
+                field_name=field,
+                field_schema=PayloadSchemaType.KEYWORD,
+                wait=True,
+            )
+            logger.info("Created KEYWORD payload index on '%s'", field)
+        except UnexpectedResponse as e:
+            # Qdrant returns 400 if the index already exists with a different
+            # schema, or simply succeeds if it already matches. Treat
+            # already-exists as benign; surface schema conflicts loudly.
+            body = getattr(e, "content", b"") or b""
+            body_text = body.decode("utf-8", errors="replace")
+            if "already exists" in body_text.lower():
+                logger.debug("Payload index on '%s' already exists", field)
+            else:
+                logger.warning(
+                    "Failed to create payload index on '%s': %s", field, body_text
+                )
+
+
+async def _has_int_doc_id_sample(
+    client: AsyncQdrantClient, collection_name: str, sample_size: int = 256
+) -> bool:
+    """Quick sample to decide whether the full backfill scroll is needed.
+
+    Reading the first batch is cheap; if all sampled doc_ids are already str
+    (the steady-state on healthy collections), we skip the full pass.
+    """
+    points, _ = await client.scroll(
+        collection_name=collection_name,
+        limit=sample_size,
+        with_payload=["doc_id"],
+        with_vectors=False,
+    )
+    for point in points:
+        payload = point.payload or {}
+        value = payload.get("doc_id")
+        if value is not None and not isinstance(value, str):
+            return True
+    return False
+
+
+async def _backfill_doc_id_to_string(
+    client: AsyncQdrantClient, collection_name: str
+) -> None:
+    """Rewrite legacy integer doc_id payloads to strings.
+
+    Producers now uniformly write str(doc_id), but historical points may carry
+    int values from before normalization. A KEYWORD index does not match int
+    payloads, so any leftover int doc_ids would be silently invisible to
+    filters. Scroll all points and convert in-place. Idempotent.
+
+    Skipped when the first sample batch already contains only str doc_ids.
+    """
+    if not await _has_int_doc_id_sample(client, collection_name):
+        logger.debug(
+            "doc_id backfill: sample shows no legacy int payloads; skipping full scan"
+        )
+        return
+
+    logger.info(
+        "Running doc_id backfill on '%s' (this may take a moment for large collections)",
+        collection_name,
+    )
+
+    rewritten = 0
+    scanned = 0
+    # Qdrant scroll returns next_offset as PointId | None — keep it untyped here
+    # so the qdrant client's full union (UUID/int/str/PointId) flows through.
+    next_offset = None
+    batch_size = 256
+
+    while True:
+        points, next_offset = await client.scroll(
+            collection_name=collection_name,
+            limit=batch_size,
+            offset=next_offset,
+            with_payload=["doc_id"],
+            with_vectors=False,
+        )
+        if not points:
+            break
+
+        for point in points:
+            scanned += 1
+            # Qdrant client typing allows None payload even when with_payload
+            # was requested; defensive default so the type checker is happy.
+            payload = point.payload or {}
+            value = payload.get("doc_id")
+            if value is None or isinstance(value, str):
+                continue
+            await client.set_payload(
+                collection_name=collection_name,
+                payload={"doc_id": str(value)},
+                points=[point.id],
+                wait=False,
+            )
+            rewritten += 1
+
+        if next_offset is None:
+            break
+
+    logger.info(
+        "doc_id backfill complete: rewrote %d/%d payloads from int to str",
+        rewritten,
+        scanned,
+    )
+
+
 async def get_qdrant_client() -> AsyncQdrantClient:
     """
     Get singleton Qdrant client instance.
@@ -110,6 +238,12 @@ async def get_qdrant_client() -> AsyncQdrantClient:
                 f"(dimension={actual_dimension}, model={settings.get_embedding_model_name()})"
             )
 
+            # Existing collections may pre-date the doc_id normalization /
+            # payload-index work. Backfill before creating the index so the
+            # index covers every point.
+            await _backfill_doc_id_to_string(_qdrant_client, collection_name)
+            await _ensure_keyword_payload_indexes(_qdrant_client, collection_name)
+
         else:
             # Collection doesn't exist - create it
             embedding_model = settings.get_embedding_model_name()
@@ -141,5 +275,6 @@ async def get_qdrant_client() -> AsyncQdrantClient:
                 f"  Distance: COSINE\n"
                 f"Background sync will index all documents with dense + sparse vectors."
             )
+            await _ensure_keyword_payload_indexes(_qdrant_client, collection_name)
 
     return _qdrant_client
diff --git a/nextcloud_mcp_server/vector/scanner.py b/nextcloud_mcp_server/vector/scanner.py
index 4f4edf3f..6a5a3758 100644
--- a/nextcloud_mcp_server/vector/scanner.py
+++ b/nextcloud_mcp_server/vector/scanner.py
@@ -48,7 +48,7 @@ class DocumentTask:
     """Document task for processing queue."""
 
     user_id: str
-    doc_id: int | str  # int for files/notes, str for legacy
+    doc_id: str  # Always str — see vector/qdrant_client.py keyword index
     doc_type: str  # "note", "file", "calendar"
     operation: str  # "index" or "delete"
     modified_at: int
@@ -228,7 +228,7 @@ async def scan_user_documents(
             )
 
             indexed_doc_ids = {
-                point.payload["doc_id"]
+                str(point.payload["doc_id"])
                 for point in (scroll_result[0] or [])
                 if point.payload is not None
             }
@@ -401,7 +401,7 @@ async def scan_user_documents(
             )
 
             indexed_file_ids = {
-                point.payload["doc_id"]
+                str(point.payload["doc_id"])
                 for point in (file_scroll_result[0] or [])
                 if point.payload is not None
             }
@@ -456,7 +456,9 @@ async def scan_user_documents(
             for file_info in tagged_files:
                 # Files are already filtered by MIME type in find_files_by_tag()
                 file_count += 1
-                file_id = file_info["id"]  # Use numeric file ID, not path
+                # Normalize file ID to str — Qdrant doc_id payload is keyword-indexed
+                # and producers across doc_types must agree on a single type.
+                file_id = str(file_info["id"])
                 file_path = file_info["path"]  # Keep path for logging
                 nextcloud_file_ids.add(file_id)
 
@@ -679,7 +681,7 @@ async def scan_news_items(
             limit=10000,
         )
         indexed_item_ids = {
-            point.payload["doc_id"]
+            str(point.payload["doc_id"])
             for point in (scroll_result[0] or [])
             if point.payload is not None
         }
@@ -858,7 +860,7 @@ async def scan_deck_cards(
             limit=10000,
         )
         indexed_card_ids = {
-            point.payload["doc_id"]
+            str(point.payload["doc_id"])
             for point in (scroll_result[0] or [])
             if point.payload is not None
         }
diff --git a/nextcloud_mcp_server/vector/visualization.py b/nextcloud_mcp_server/vector/visualization.py
index 5ffb4bd1..3da8d2e6 100644
--- a/nextcloud_mcp_server/vector/visualization.py
+++ b/nextcloud_mcp_server/vector/visualization.py
@@ -70,7 +70,10 @@ async def compute_pca_coordinates(
                 vector = point.vector
 
             if vector is not None and point.payload:
-                doc_id = point.payload.get("doc_id")
+                # SearchResult.id is str; coerce payload doc_id to match so the
+                # tuple lookup below succeeds even on legacy int-typed payloads.
+                raw_doc_id = point.payload.get("doc_id")
+                doc_id = None if raw_doc_id is None else str(raw_doc_id)
                 chunk_start = point.payload.get("chunk_start_offset")
                 chunk_end = point.payload.get("chunk_end_offset")
                 chunk_key = (doc_id, chunk_start, chunk_end)
diff --git a/tests/unit/vector/__init__.py b/tests/unit/vector/__init__.py
new file mode 100644
index 00000000..e69de29b
diff --git a/tests/unit/vector/test_qdrant_client.py b/tests/unit/vector/test_qdrant_client.py
new file mode 100644
index 00000000..db232ddd
--- /dev/null
+++ b/tests/unit/vector/test_qdrant_client.py
@@ -0,0 +1,265 @@
+"""Unit tests for Qdrant payload-index helpers and doc_id backfill.
+
+These cover the startup-time migrations added to ``vector/qdrant_client.py``
+after production HTTP 400 errors revealed that:
+
+1. The collection had no payload index for ``doc_id``, so any
+   ``FieldCondition(key="doc_id", ...)`` filter failed at the Qdrant layer.
+2. Producers wrote a mix of ``int`` and ``str`` values for ``doc_id``, so a
+   single keyword index could not have covered both kinds even if it had
+   existed.
+
+The fix has three coordinated parts; this module covers the two helpers that
+run at startup. Producer-side normalization is exercised by the existing
+scanner tests.
+"""
+
+from types import SimpleNamespace
+from unittest.mock import call
+
+import httpx
+import pytest
+from qdrant_client.http.exceptions import UnexpectedResponse
+from qdrant_client.models import PayloadSchemaType
+
+from nextcloud_mcp_server.vector.qdrant_client import (
+    _KEYWORD_PAYLOAD_FIELDS,
+    _backfill_doc_id_to_string,
+    _ensure_keyword_payload_indexes,
+    _has_int_doc_id_sample,
+)
+
+
+def _make_unexpected(status_code: int, body: bytes) -> UnexpectedResponse:
+    """Build a real UnexpectedResponse for raise_for_status-style branches."""
+    return UnexpectedResponse(
+        status_code=status_code,
+        reason_phrase="Bad Request",
+        content=body,
+        headers=httpx.Headers(),
+    )
+
+
+def _record(point_id: int | str, doc_id: int | str | None) -> SimpleNamespace:
+    """Stand-in for qdrant_client.http.models.Record.
+
+    Tests don't need full Pydantic validation — only ``id`` and ``payload``
+    are read by the helpers under test.
+    """
+    payload: dict | None = {"doc_id": doc_id} if doc_id is not None else None
+    return SimpleNamespace(id=point_id, payload=payload)
+
+
+# ---------------------------------------------------------------------------
+# _ensure_keyword_payload_indexes
+# ---------------------------------------------------------------------------
+
+
+@pytest.mark.unit
+async def test_ensure_keyword_payload_indexes_creates_each_field(mocker):
+    """Happy path: every field in _KEYWORD_PAYLOAD_FIELDS gets a KEYWORD index."""
+    client = mocker.AsyncMock()
+
+    await _ensure_keyword_payload_indexes(client, "test-collection")
+
+    assert client.create_payload_index.await_count == len(_KEYWORD_PAYLOAD_FIELDS)
+    expected_calls = [
+        call(
+            collection_name="test-collection",
+            field_name=field,
+            field_schema=PayloadSchemaType.KEYWORD,
+            wait=True,
+        )
+        for field in _KEYWORD_PAYLOAD_FIELDS
+    ]
+    client.create_payload_index.assert_has_awaits(expected_calls, any_order=False)
+
+
+@pytest.mark.unit
+async def test_ensure_keyword_payload_indexes_swallows_already_exists(mocker, caplog):
+    """Idempotent: 'already exists' 400 is logged at debug, not raised."""
+    client = mocker.AsyncMock()
+    # First call succeeds, second raises "already exists", third succeeds —
+    # exercises the per-field exception handling.
+    client.create_payload_index.side_effect = [
+        None,
+        _make_unexpected(
+            400, b'{"status":{"error":"Index for \\"user_id\\" already exists"}}'
+        ),
+        None,
+    ]
+
+    with caplog.at_level("DEBUG", logger="nextcloud_mcp_server.vector.qdrant_client"):
+        await _ensure_keyword_payload_indexes(client, "test-collection")
+
+    assert client.create_payload_index.await_count == len(_KEYWORD_PAYLOAD_FIELDS)
+    # The "already exists" branch logs at DEBUG; nothing reaches WARNING.
+    assert not any(record.levelname == "WARNING" for record in caplog.records)
+
+
+@pytest.mark.unit
+async def test_ensure_keyword_payload_indexes_logs_unrelated_400_as_warning(
+    mocker, caplog
+):
+    """Schema conflicts and other 400s are surfaced as warnings, not silenced."""
+    client = mocker.AsyncMock()
+    client.create_payload_index.side_effect = [
+        _make_unexpected(
+            400,
+            b'{"status":{"error":"field \\"doc_id\\" indexed with different schema"}}',
+        ),
+        None,
+        None,
+    ]
+
+    with caplog.at_level("WARNING", logger="nextcloud_mcp_server.vector.qdrant_client"):
+        await _ensure_keyword_payload_indexes(client, "test-collection")
+
+    # Loop continued past the failing field; all three were attempted.
+    assert client.create_payload_index.await_count == len(_KEYWORD_PAYLOAD_FIELDS)
+    warnings = [r for r in caplog.records if r.levelname == "WARNING"]
+    assert len(warnings) == 1
+    assert "different schema" in warnings[0].getMessage()
+
+
+# ---------------------------------------------------------------------------
+# _has_int_doc_id_sample
+# ---------------------------------------------------------------------------
+
+
+@pytest.mark.unit
+async def test_has_int_doc_id_sample_returns_true_when_int_present(mocker):
+    """Sample finds an int — caller should run the full backfill."""
+    client = mocker.AsyncMock()
+    client.scroll.return_value = (
+        [_record(1, "abc"), _record(2, 42), _record(3, "xyz")],
+        None,
+    )
+
+    assert await _has_int_doc_id_sample(client, "c") is True
+    client.scroll.assert_awaited_once()
+
+
+@pytest.mark.unit
+async def test_has_int_doc_id_sample_returns_false_when_all_str(mocker):
+    """Sample is clean — caller should skip the full scroll."""
+    client = mocker.AsyncMock()
+    client.scroll.return_value = (
+        [_record(1, "abc"), _record(2, "def")],
+        None,
+    )
+
+    assert await _has_int_doc_id_sample(client, "c") is False
+
+
+@pytest.mark.unit
+async def test_has_int_doc_id_sample_handles_empty_collection(mocker):
+    """Empty collection — nothing to backfill, return False."""
+    client = mocker.AsyncMock()
+    client.scroll.return_value = ([], None)
+
+    assert await _has_int_doc_id_sample(client, "c") is False
+
+
+@pytest.mark.unit
+async def test_has_int_doc_id_sample_ignores_missing_payload(mocker):
+    """Records with no payload don't count as int doc_ids."""
+    client = mocker.AsyncMock()
+    client.scroll.return_value = (
+        [_record(1, None), _record(2, "abc")],
+        None,
+    )
+
+    assert await _has_int_doc_id_sample(client, "c") is False
+
+
+# ---------------------------------------------------------------------------
+# _backfill_doc_id_to_string
+# ---------------------------------------------------------------------------
+
+
+@pytest.mark.unit
+async def test_backfill_skips_when_sample_is_clean(mocker):
+    """Short-circuit: clean sample → no full scroll, no set_payload calls."""
+    client = mocker.AsyncMock()
+    # Sample call returns only str payloads → backfill should not proceed.
+    client.scroll.return_value = ([_record(1, "abc"), _record(2, "def")], None)
+
+    await _backfill_doc_id_to_string(client, "test-collection")
+
+    # Exactly one scroll (the sample) and zero rewrites.
+    assert client.scroll.await_count == 1
+    client.set_payload.assert_not_awaited()
+
+
+@pytest.mark.unit
+async def test_backfill_rewrites_int_doc_ids_to_str(mocker):
+    """Mixed int/str payload across two scroll pages: only ints get rewritten."""
+    client = mocker.AsyncMock()
+    # Three scroll calls:
+    #   1. sample → finds an int, triggers the full pass
+    #   2. first batch of full scroll → mixed int/str
+    #   3. second batch → all str, with next_offset=None to terminate
+    client.scroll.side_effect = [
+        ([_record(1, 100), _record(2, "abc")], None),  # sample
+        ([_record(1, 100), _record(2, "abc")], "next-offset-123"),  # batch 1
+        ([_record(3, 200), _record(4, "def")], None),  # batch 2 (terminal)
+    ]
+
+    await _backfill_doc_id_to_string(client, "test-collection")
+
+    # Two rewrites: point 1 (int 100) in batch 1, point 3 (int 200) in batch 2.
+    assert client.set_payload.await_count == 2
+    client.set_payload.assert_any_await(
+        collection_name="test-collection",
+        payload={"doc_id": "100"},
+        points=[1],
+        wait=False,
+    )
+    client.set_payload.assert_any_await(
+        collection_name="test-collection",
+        payload={"doc_id": "200"},
+        points=[3],
+        wait=False,
+    )
+
+
+@pytest.mark.unit
+async def test_backfill_emits_completion_log(mocker, caplog):
+    """Backfill logs final rewritten/scanned counts at INFO."""
+    client = mocker.AsyncMock()
+    client.scroll.side_effect = [
+        ([_record(1, 7)], None),  # sample triggers full pass
+        ([_record(1, 7), _record(2, "x")], None),  # single batch, terminal
+    ]
+
+    with caplog.at_level("INFO", logger="nextcloud_mcp_server.vector.qdrant_client"):
+        await _backfill_doc_id_to_string(client, "test-collection")
+
+    completion_logs = [
+        r.getMessage() for r in caplog.records if "backfill complete" in r.getMessage()
+    ]
+    assert completion_logs, "expected an INFO log line for backfill completion"
+    msg = completion_logs[0]
+    assert "1/2" in msg, f"expected '1/2' rewritten/scanned in {msg!r}"
+
+
+@pytest.mark.unit
+async def test_backfill_handles_none_payload(mocker):
+    """A point with payload=None is skipped without crashing."""
+    client = mocker.AsyncMock()
+    client.scroll.side_effect = [
+        ([_record(1, 99)], None),  # sample triggers full pass
+        ([_record(1, None), _record(2, 99)], None),  # batch with one None payload
+    ]
+
+    await _backfill_doc_id_to_string(client, "test-collection")
+
+    # Only the int doc_id at point 2 was rewritten; the None-payload point was skipped.
+    assert client.set_payload.await_count == 1
+    client.set_payload.assert_awaited_with(
+        collection_name="test-collection",
+        payload={"doc_id": "99"},
+        points=[2],
+        wait=False,
+    )