Add generic LLM worker

k8s/configmap.yaml

@@ -12,3 +12,5 @@ data:
   LLM_MODEL: "qwen2.5-14b"
   LLM_TIMEOUT: "5m"
   WHISPERX_URL: "http://10.2.3.5:8001"
+  WORKER_POLL_INTERVAL: "2s"
+  WORKER_CLAIM_LIMIT: "4"

k8s/kustomization.yaml

@@ -10,3 +10,4 @@ resources:
   - postgres.yaml
   - server-deployment.yaml
   - server-service.yaml
+  - worker-deployment.yaml

k8s/worker-deployment.yaml

@@ -0,0 +1,37 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: ai-service-worker
+  namespace: ai-service
+spec:
+  replicas: 1
+  selector:
+    matchLabels:
+      app: ai-service-worker
+  template:
+    metadata:
+      labels:
+        app: ai-service-worker
+    spec:
+      terminationGracePeriodSeconds: 20
+      containers:
+        - name: worker
+          image: localhost:30300/admin/ai-service:latest
+          command: ["/usr/local/bin/ai-service-worker"]
+          env:
+            - name: WORKER_ID
+              valueFrom:
+                fieldRef:
+                  fieldPath: metadata.name
+          envFrom:
+            - configMapRef:
+                name: ai-service-config
+            - secretRef:
+                name: ai-service-secrets
+          resources:
+            requests:
+              cpu: 50m
+              memory: 96Mi
+            limits:
+              cpu: 500m
+              memory: 384Mi