Kotlin Jupyter Kernel with Vertex AI Workbench

This container can be used as a custom container for Google Cloud Vertex AI Workbench instances.

The architecture to build a Vertex AI workbench container image.

Lab 0: Create Vertex AI Workbench

Step 1: Select the project and open Google Cloud Shell

Login to Google Cloud Platform. Select the right project and click to open the Google Cloud Shell as shown below.

Step 2: Create Kotlin Jupyter Vertex AI Workbench (takes 10 minutes)

Accept and authorize to use Google Cloud Shell. Copy and paste the following scripts. The script will do the following:

2.1 Enable all the services in Google Cloud

2.2 Create Artifacts Repository for Kotlin Jupyter container image.

2.3 Create serverless Cloud Build to build the image from this repository's Dockerfile. See the cloudbuild.yaml configuration.

2.4 Create Vertex AI workbench instance

# clone the repository
echo "[Step 2] Clone the repository..."
cd ~
git clone https://github.com/gaplo917/kotlin-vertexai-workbench-container.git || true
cd kotlin-vertexai-workbench-container
git fetch --all && git reset --hard origin/main

# enable services APIs
echo "[Step 2.1] Enable services APIs..."
gcloud services enable aiplatform.googleapis.com
gcloud services enable artifactregistry.googleapis.com
gcloud services enable cloudbuild.googleapis.com
gcloud services enable containerfilesystem.googleapis.com
gcloud services enable containerregistry.googleapis.com


PROJECT_ID=$(gcloud config get-value project)

# create artifacts repository
echo "[Step 2.2] Create artifacts repository..."
gcloud artifacts repositories create demo --repository-format=docker --location=us --project=$PROJECT_ID || true

# create build to push image to repository
echo "[Step 2.3] create build to push image to repository..."
gcloud builds submit --config cloudbuild.yaml

# create Vertex AI workbench instance
echo "[Step 2.4] create Vertex AI workbench instance..."
gcloud workbench instances create kotlin-workbench-instance \
  --metadata=idle-timeout-seconds=10800 \
  --location=us-central1-a \
  --container-repository=us-docker.pkg.dev/$PROJECT_ID/demo/kotlin-vertexai-workbench-container \
  --container-tag=latest \
  --machine-type=e2-standard-4 \
  --project=$PROJECT_ID
  
echo "[Completed] You can go to https://console.cloud.google.com/vertex-ai/workbench/instances"

Step 3: Go to Vertex AI workbench

https://console.cloud.google.com/vertex-ai/workbench/instances

Lab 1: Run Kotlin Jupyter Notebook for Gemini

Step 1: Open Jupyter Lab > Terminal

Step 2: Clone the repository

• Type git clone https://github.com/gaplo917/awesome-kotlin-notebook/ for Lab 1, 3
• Type git clone https://github.com/gaplo917/kotlin-cyrpto-price-spring-mcp-server-demo for Lab 3

Step 3: Get your Project ID

Type echo $(gcloud config get-value project) to get your project ID

Step 4: Open the Kotlin Jupyter Notebook

Navigate awesome-kotlin-notebook/google-genai-sdk/ in the file browser. Open vertexai-gemini-examples.ipynb.

Step 5: Run the Kotlin Notebook

Remember to replace your project ID in one of the cells

Lab 2: Run Spring Boot MCP Server

• Navigate kotlin-cyrpto-price-spring-mcp-server-demo/ in the file browser
• Open notebook.ipynb
• Start Spring Boot MCP Server and proceed to Lab 3

The MCP Architecture

Lab 3: Run Kotlin Jupyter Notebook for Langchain4j MCP Client

Follow up on Lab 1:

• Navigate awesome-kotlin-notebook/langchain4j/ in the file browser
• Open langchain4j-mcp-client-vertexai.ipynb
• Run the Kotlin Notebook, remember to replace your project ID in one of the cells

Completed all labs

Congratulation! You have completed all the labs! Remember to delete Vertex AI instance to avoid unnecessary charges.