📌 Part 3: Creating an HTTP API in AWS API Gateway to Access Lambda (with Screenshot Guides)

1. Create an HTTP API via API Gateway

Purpose: The HTTP API allows you to expose the Lambda function via a public HTTP endpoint that you can call from external applications (e.g., Postman, Jupyter Notebook).

Steps:

• Go to the AWS Console → API Gateway → APIs.
• Click on Create API.
• Choose HTTP API and click Build.

🔹 Screenshot Tip:
🔹 Screenshot of the API Gateway landing page → Create API

2. Configure API Integration with Lambda

Steps:

• Add Integration:
  • Integration type: Lambda function.
  • Choose your Lambda function: generate-text-lamini.
  • Check the box Add permissions to Lambda function — this allows API Gateway to invoke your Lambda automatically.
• Click Next.

🔹 Screenshot Tip:
🔹 Screenshot the Add Integration screen showing your Lambda function selected

3. Configure Route

Steps:

• Define Route:
  • Method: POST
  • Resource path: /generate
• Click Next.

🔹 Screenshot Tip:
🔹 Screenshot the Route configuration screen showing POST and /generate

4. Configure Stage

Steps:

• Stage name: leave as $default (or set a custom stage name if preferred).
• Enable auto-deploy (this publishes changes automatically without extra steps).
• Click Next → Create.

🔹 Screenshot Tip:
🔹 Screenshot the Stage configuration screen with $default and Auto deploy enabled

5. Find Your Invoke URL

Steps:

• After creation, you will land on the API dashboard.
• You’ll see the Invoke URL at the top.
  Format example:

https://your-api-id.execute-api.us-east-2.amazonaws.com

• Append /generate to the Invoke URL.
  🔹 Final API Endpoint:

https://your-api-id.execute-api.us-east-2.amazonaws.com/generate

🔹 Screenshot Tip:
🔹 Screenshot of the API Invoke URL and the Routes tab showing /generate

6. (Optional) Test API Directly in AWS Console

• Go to API Gateway → APIs → your API → Routes → POST /generate → Test.
• Request Body:

{
  "prompt": "Write a poem about the ocean."
}

• Click Test and verify the Lambda/SageMaker pipeline works.

🔹 Screenshot Tip:
🔹 Screenshot of the API Gateway testing console showing request/response

7. Example API Request from External Client (Python)

You can now interact with your model remotely using a simple Python script.

import requests

# Replace with your actual invoke URL
url = "https://your-api-id.execute-api.us-east-2.amazonaws.com/generate"

data = {"prompt": "Write an article about deploying LLM models to AWS services"}

response = requests.post(url, json=data)

print(response.json())

🔹 Screenshot Tip:
🔹 Optional: Screenshot of API call from a Jupyter Notebook or terminal

🔹 Summary Checklist:

• ✅ Created HTTP API with Lambda integration.
• ✅ Configured POST /generate route linked to Lambda.
• ✅ Auto-deployed to $default stage.
• ✅ Retrieved and validated public HTTP endpoint.
• ✅ Tested end-to-end pipeline from API Gateway → Lambda → SageMaker Endpoint.

🔹 Important Notes:

• ✅ Ensure your Lambda function’s IAM role includes sagemaker:InvokeEndpoint permission (covered in Part 2).
• ✅ The Invoke URL is what you will use to send prompts to your model.
• ✅ You can monitor Lambda/API logs in CloudWatch Logs.

🔹 Next Step: You are now ready to send HTTP requests to your SageMaker-powered LLM model through API Gateway!