Gen-AI Developer Classroom notes 29/Mar/2026

Configuring Models from AWS Bedrock

• Ensure all the necessary softwares are installed

winget install --id Git.Git -e --source winget
winget install -e --id Python.Python.3.13
winget install -e --id astral-sh.uv
winget install --id Microsoft.VisualStudioCode -e --source winget
winget install --id Amazon.AWSCLI -e --source winget
winget install Microsoft.AzureCLI

• Create an IAM User (Watch Classroom recording)
• Use North Virgina or us-east-1 as region for Bedrock
• Now configure aws with access key and secret key aws configure
• Enable bedrock access AWS Console -> BedRoc -> Model Access -> Request Access

Now lets create a sample to verify aws bedrock access

• Refer Here for langchains documentation

• Create a new directory

mkdir aws_connectivity_test
cd aws_connectivity_test
uv init
uv sync
uv add langchain langchain-aws python-dotenv

• Pyproject.toml

[project]
name = "aws-test"
version = "0.1.0"
description = "Add your description here"
readme = "README.md"
requires-python = ">=3.13"
dependencies = [
    "langchain>=1.2.13",
    "langchain-aws>=1.4.1",
    "python-dotenv>=1.2.2",
]

• main.py

from langchain_aws import ChatBedrockConverse


def main():
    llm  = ChatBedrockConverse(
        model="amazon.nova-micro-v1:0",
        region_name="us-east-1",
        temperature=0,
        max_tokens=256
    )
    # Ask a question
    response = llm.invoke("What is the capital of France?")

    print(response.content)


if __name__ == "__main__":
    main()

• Now run uv run main.py to get the response

Configuring Models from Azure AI Foundry

• Ensure Azure CLI is installed
• Configure Azure login

az login

• Enabling Azure AI Foundry for model access

how to enable azure ai foundry for serverless models
I want to cheaper phi models with tool calling

• Refer Here for using langchain or langgraph with azure ai

• After enabling a model, Create a new folder

mkdir azure_test
cd azure_test
uv init
uv sync
uv add langchain langchain-azure-ai azure-identity python-dotenv

• Todo: Need to configure correctly