feat: LLM routing by tier (free→Ollama, pro→Timeweb)

- Add tier-based provider routing in llm-svc
  - free tier → Ollama (local qwen3.5:9b)
  - pro/business → Timeweb Cloud AI
- Add /api/v1/embed endpoint for embeddings via Ollama
- Update Ollama client: qwen3.5:9b default, remove auth
- Add GenerateEmbedding() function for qwen3-embedding:0.6b
- Add Ollama K8s deployment with GPU support (RTX 4060 Ti)
- Add monitoring stack (Prometheus, Grafana, Alertmanager)
- Add Grafana dashboards for LLM and security metrics
- Update deploy.sh with monitoring and Ollama deployment

Made-with: Cursor

backend/deploy/k8s/ollama-models.yaml

@@ -0,0 +1,85 @@
+# Job для загрузки моделей Ollama после деплоя
+apiVersion: batch/v1
+kind: Job
+metadata:
+  name: ollama-model-loader
+  namespace: gooseek
+  labels:
+    app: ollama-model-loader
+spec:
+  ttlSecondsAfterFinished: 3600
+  backoffLimit: 3
+  template:
+    metadata:
+      labels:
+        app: ollama-model-loader
+    spec:
+      restartPolicy: OnFailure
+      initContainers:
+      - name: wait-for-ollama
+        image: curlimages/curl:latest
+        command:
+        - /bin/sh
+        - -c
+        - |
+          echo "Waiting for Ollama to be ready..."
+          until curl -sf http://ollama.gooseek.svc.cluster.local:11434/api/tags; do
+            echo "Ollama not ready, retrying in 5s..."
+            sleep 5
+          done
+          echo "Ollama is ready!"
+      containers:
+      - name: model-loader
+        image: ollama/ollama:latest
+        env:
+        - name: OLLAMA_HOST
+          value: "http://ollama.gooseek.svc.cluster.local:11434"
+        command:
+        - /bin/sh
+        - -c
+        - |
+          set -e
+          
+          OLLAMA_URL="http://ollama.gooseek.svc.cluster.local:11434"
+          
+          pull_model() {
+            MODEL=$1
+            echo "=== Pulling model: $MODEL ==="
+            
+            # Check if model already exists
+            EXISTING=$(curl -sf "$OLLAMA_URL/api/tags" | grep -o "\"name\":\"$MODEL\"" || true)
+            if [ -n "$EXISTING" ]; then
+              echo "Model $MODEL already exists, skipping..."
+              return 0
+            fi
+            
+            # Pull model via API
+            echo "Downloading $MODEL..."
+            curl -sf "$OLLAMA_URL/api/pull" \
+              -H "Content-Type: application/json" \
+              -d "{\"name\": \"$MODEL\", \"stream\": false}" \
+              --max-time 1800
+            
+            echo "Model $MODEL downloaded successfully!"
+          }
+          
+          echo "=== Ollama Model Loader ==="
+          echo "Target: $OLLAMA_URL"
+          
+          # Основная модель генерации (4 параллельных воркера)
+          pull_model "qwen3.5:9b"
+          
+          # Embedding модель (быстрые эмбеддинги)
+          pull_model "qwen3-embedding:0.6b"
+          
+          echo ""
+          echo "=== All models loaded ==="
+          curl -sf "$OLLAMA_URL/api/tags" | head -c 500
+          echo ""
+        resources:
+          requests:
+            cpu: 100m
+            memory: 256Mi
+          limits:
+            cpu: 500m
+            memory: 512Mi