This project contains my first agent built using Google ADK (Agent Development Kit) while following the official ADK course. It's a learning exercise to understand how to create and deploy AI agents.
The agent is designed as a Math Tutor that assists students with algebra problems. It leverages Google's Gemini 2.5 Flash model to provide patient, step-by-step guidance for problem-solving.
- Name: math_tutor_agent
- Model: gemini-2.5-flash
- Purpose: Help students learn algebra by guiding them through problem-solving steps
- Personality: Patient math tutor
- Python 3.11 or higher
- Google Cloud account with ADK access
- Git
-
Clone the repository:
git clone <your-repo-url> cd adk-workspace
-
Create a virtual environment:
python -m venv .venv
-
Activate the virtual environment:
- On Windows:
& .venv\Scripts\Activate.ps1 - On macOS/Linux:
source .venv/bin/activate
- On Windows:
-
Install dependencies:
pip install google-adk
-
Set up Google Cloud credentials (if required for ADK):
- Follow the ADK documentation for authentication
-
Run the agent locally:
adk run my_first_agent
-
Interact with the agent:
- The agent will respond to algebra-related queries
- Example: "Help me solve 2x + 3 = 7"
adk-workspace/
├── .adk/ # ADK configuration files
├── my_first_agent/ # Agent package
│ ├── __init__.py # Package initialization
│ └── agent.py # Agent definition
├── .gitignore # Git ignore rules
├── .venv/ # Virtual environment (ignored)
└── README.md # This file
This project is part of my ADK learning path. Key concepts covered:
- Agent creation with Google ADK
- Model selection (Gemini 2.5 Flash)
- Agent configuration (name, description, instructions)
- Basic agent deployment
- Add more complex agent behaviors
- Implement multi-turn conversations
- Add evaluation metrics
- Deploy to Google Cloud
This is a personal learning project, but feel free to suggest improvements!
This project is for educational purposes.