Review the shopping assistant implementation

Inspect the assistant files

The source tree already contains the missing AI tier. Before you build it, review the files that make up the assistant:

    

        
        
find src/shoppingassistantservice -maxdepth 2 -type f | sort
find kustomize/components/shopping-assistant -maxdepth 2 -type f | sort

    

You’ll see the assistant source folder and the shopping-assistant Kustomize component.

Review the runtime dependencies and gRPC stubs

The assistant uses a small Python runtime and generated gRPC stubs to call the live storefront services:

    

        
        
sed -n '1,20p' src/shoppingassistantservice/requirements.txt
sed -n '1,25p' src/shoppingassistantservice/demo_pb2.py
grep -nE 'class (CartServiceStub|ProductCatalogServiceStub)' \
  src/shoppingassistantservice/demo_pb2_grpc.py

    

The generated protobuf files are dense because they’re machine-generated. You only need to confirm that the cart and product catalog stubs exist.

Review the assistant service

Read the beginning and end of the assistant service:

    

        
        
sed -n '1,220p' src/shoppingassistantservice/shoppingassistantservice.py
tail -n 20 src/shoppingassistantservice/shoppingassistantservice.py

    

Focus on these parts of the file:

• Imports and environment variables
• Tool functions that call the live storefront services
• Request handlers that decide how the assistant responds
• The app.run(...) entrypoint

Highlight the tool functions:

    

        
        
grep -nE 'def (search_catalog|get_product_details|get_cart|add_to_cart|handle_[a-z_]+)' \
  src/shoppingassistantservice/shoppingassistantservice.py

    

These functions define the assistant behavior. The service searches the catalog, fetches product details, reads the cart, and updates the cart after confirmation. These are live application tool calls, not mock responses, which is what makes the assistant part of the storefront workflow.

Confirm that the Python source is valid

Compile the assistant source:

    

        
        
python3 -m py_compile src/shoppingassistantservice/shoppingassistantservice.py

    

The command returns no output. Don’t continue if this command fails.

Review the container and Kubernetes component

Review the Dockerfile and the Kustomize component:

    

        
        
sed -n '1,80p' src/shoppingassistantservice/Dockerfile
sed -n '1,120p' kustomize/components/shopping-assistant/kustomization.yaml
sed -n '1,220p' kustomize/components/shopping-assistant/shoppingassistantservice.yaml

    

The Dockerfile packages the Python assistant service. The Kubernetes component creates shoppingassistantservice, adds the Ollama sidecar, and patches frontend so the assistant becomes part of the storefront experience. Because the assistant and Ollama run in the same pod, moving this deployment later moves the AI logic and local reasoning runtime together.

What you’ve accomplished and what’s next

You’ve now reviewed the source code, service contracts, container packaging, and Kubernetes component for the assistant tier.

Next, you’ll build the assistant image for linux/arm64 and push it to Artifact Registry.