full ingestion -> OCR -> extraction flow is now working correctly.

apps/svc_ocr/Dockerfile

@@ -21,8 +21,10 @@ RUN apt-get update && apt-get install -y \
 WORKDIR /app
 
 # Copy service-specific requirements and install
+# Copy base requirements and service-specific requirements
+COPY libs/requirements-base.txt /tmp/libs-requirements.txt
 COPY apps/svc_ocr/requirements.txt /tmp/service-requirements.txt
-RUN pip install --no-cache-dir -r /tmp/service-requirements.txt
+RUN pip install --no-cache-dir -r /tmp/libs-requirements.txt -r /tmp/service-requirements.txt
 
 # Copy application code
 COPY libs/ ./libs/

apps/svc_ocr/main.py

@@ -118,7 +118,7 @@ async def init_dependencies(app_settings: OCRSettings) -> None:
             if attempt == max_retries:
                 raise HTTPException(
                     status_code=500, detail="Failed to connect to NATS after retries"
-                )
+                ) from e
             await asyncio.sleep(delay)
             delay *= 2  # exponential backoff
 
@@ -280,7 +280,7 @@ async def _handle_document_ingested(topic: str, payload: EventPayload) -> None:
         return
 
     # Auto-process PDF documents
-    if data.get("content_type") == "application/pdf":
+    if data.get("mime_type") == "application/pdf":
         logger.info("Auto-processing ingested document", doc_id=doc_id)
 
         try:
@@ -347,13 +347,13 @@ async def _process_document_async(
             await ds.store_ocr_result(tenant_id, doc_id, ocr_results)
 
             # Update metrics
-            metrics.counter("documents_processed_total").labels(
-                tenant_id=tenant_id, strategy=strategy
-            ).inc()
+            metrics.counter(
+                "ocr_documents_processed_total", labelnames=["tenant_id", "strategy"]
+            ).labels(tenant_id=tenant_id, strategy=strategy).inc()
 
-            metrics.histogram("processing_duration_seconds").labels(
-                strategy=strategy
-            ).observe(
+            metrics.histogram(
+                "ocr_processing_duration_seconds", labelnames=["strategy"]
+            ).labels(strategy=strategy).observe(
                 datetime.utcnow().timestamp()
                 - datetime.fromisoformat(
                     ocr_results["processed_at"].replace("Z", "")  # type: ignore
@@ -386,7 +386,10 @@ async def _process_document_async(
             logger.error("OCR processing failed", doc_id=doc_id, error=str(e))
 
             # Update error metrics
-            metrics.counter("processing_errors_total").labels(
+            metrics.counter(
+                "ocr_processing_errors_total",
+                labelnames=["tenant_id", "strategy", "error_type"],
+            ).labels(
                 tenant_id=tenant_id, strategy=strategy, error_type=type(e).__name__
             ).inc()