Azure AI Agent Workshop

2025 TrendMicro AI Hackathon x Microsoft Azure AI Agent Workshop

Clone the repository

git clone https://github.com/LittleFish-Coder/azure-ai-agent-workshop.git

1. Prerequisite

Please refer to Prerequisite.md for detailed instructions.

In the prerequisite step, you'll complete:

1. Set up your Python 3.11+ environment (install Python, create & activate a virtual environment, update pip, install dependencies).
2. Install and configure the Azure CLI (install, check version, login, select subscription).

2. Setup Azure Service Environment

Modify .env file with your Azure resource information.

# Azure AI Foundry Agent Service
AZURE_AI_AGENT_ENDPOINT="https://<example-project-name>.services.ai.azure.com/api/projects/<example-project-name>"
AZURE_AI_AGENT_MODEL_DEPLOYMENT_NAME="gpt-4o"

# Azure Bing Search Service
AZURE_BING_CONNECTION_ID=""

# Azure AI Search Service
AZURE_SEARCH_ENDPOINT="https://<example-search-name>.search.windows.net"
AZURE_SEARCH_API_KEY=""

# Azure OpenAI Service
AZURE_OPENAI_ENDPOINT="https://<example-project-name>.openai.azure.com/openai/deployments/<example-model-deployment-name>/chat/completions?api-version=<example-api-version>"
AZURE_OPENAI_API_KEY="<example-api-key>"
AZURE_OPENAI_API_VERSION="<example-api-version>"
AZURE_OPENAI_CHAT_DEPLOYMENT_NAME="gpt-4o"
AZURE_OPENAI_GPT_MODEL="gpt-4o"         # for s2_ai_search
AZURE_OPENAI_GPT_DEPLOYMENT="gpt-4o"    # for s2_ai_search

# AZURE OpenAI Embedding Model Service
AZURE_OPENAI_EMBEDDING_MODEL="text-embedding-3-large"       # for s2_ai_search
AZURE_OPENAI_EMBEDDING_DEPLOYMENT="text-embedding-3-large"  # for s2_ai_search

3. Run the Lab

We will go through the following labs:

• Single Agent: s1_single_agent/
• AI Search: s2_ai_search/
• Multi-Agent: s3_multi_agent/

1. Activate the virtual environment

.venv\Scripts\activate     # Windows PowerShell

source .venv/bin/activate   # Mac/Linux

2. Test the environment with the following command:

python s3_multi_agent/0_setup/test_azure_ai_agent.py

python s3_multi_agent/0_setup/test_chat_completion.py

python s3_multi_agent/0_setup/test_multi_agent.py

3. Run the lab with the following command:

python s1_single_agent/lab_name.py

# ... etc