Jarvis

Project Folder Structure

ADK_Hackathon/
├── backend/                
│   ├── personal_assistant/ # Python FastAPI backend with sub-agents and utilities
│   │   ├── agent.py # root_agent Jarvis
│   │   ├── app.py
│   │   ├── prompt.py # root_agent prompt
│   │   ├── email_agent/ # email_agent setup
│   │   ├── calendar_agent/ # calendar_agent setup
│   │   ├── mobility_agent/ # mobility_agent setup
│   │   ├── common/ # common utils
│   │   │   ├── __init__.py
│   │   │   ├── agent_utils.py
│   │   │   ├── tool_utils.py
│   ├── deployment  # deployyment using google vertexai
│   ├── .env  # enviroment variables
│   ├── poetry.lock 
│   └── pyproject.toml      # Python dependencies for Poetry
├── frontend/               # React frontend application
│   ├── index.html
│   ├── index.tsx
│   └── ...
├── README.md               # Project documentation (this file)

This contains everything you need to run your app locally.

Run Backend Locally

Prerequisites:

• python 3.13.2 (recommend use pyenv for version management)
• poetry (please follow this link: https://python-poetry.org/docs/)
• have to setup google api project (https://console.cloud.google.com/) properly and add all the APIs ( "Compute Engine API", "Vertex AI API", "Cloud Logging API", "Geocoding API", "Weather API", "Places API (New)", "Gemini for Google Cloud API", "Google Calendar API" )
• store all credentials in your .env

1. Install dependencies:

cd backend
poetry install

2. run the fastapi app:

cd personal_assistant
poetry run uvicorn app:app --reload

Run Frontend Locally

Prerequisites: Node.js

1. Install dependencies: npm install
2. Run the app: npm run dev

You can now play with the UI using the link provided by npm, which connects to your fastapi locally