ASCA - Adaptive Student Coaching Agents

Multi-agent AI system helping students manage academic workload and wellness through intelligent collaboration.

🎯 Overview

ASCA uses three specialized AI agents powered by Google Gemini Pro that communicate with each other to provide comprehensive academic support:

1. Assignment Analyzer Agent - Analyzes workload complexity and estimates time requirements
2. Schedule Optimizer Agent - Creates optimized study schedules with breaks
3. Wellness Monitor Agent - Assesses student wellness and provides recommendations

🤖 Multi-Agent Architecture

Student Input → Assignment Analyzer → Schedule Optimizer → Wellness Monitor → Complete Analysis
                      ↓                      ↓                      ↓
                 Workload Data         Schedule Data         Wellness Data

Each agent sends structured JSON messages to the next, creating true multi-agent collaboration.

🚀 Quick Start

Prerequisites

• Python 3.11+
• Google Gemini API key

Installation

cd backend
pip install -r requirements.txt

Run the Demo

# Test the multi-agent system
python test_agents.py

# Start the API server
python main.py

Visit http://localhost:8080/docs for API documentation.

Run the Frontend

cd ..
python -m http.server 8000

Visit http://localhost:8000 for the web interface.

📡 API Endpoints

• POST /api/analyze-assignment - Analyze single assignment
• POST /api/analyze-workload - Analyze multiple assignments
• POST /api/create-schedule - Create optimized schedule (2 agents)
• POST /api/wellness-check - Wellness assessment (3 agents)
• POST /api/full-analysis - Complete multi-agent workflow

🏗️ Tech Stack

• AI: Google Gemini Pro API
• Backend: FastAPI (Python 3.11)
• Deployment: Google Cloud Run
• Frontend: HTML/CSS/JavaScript

🌐 Deployment

Deploy to Google Cloud Run:

cd backend
gcloud run deploy asca-backend \
  --source . \
  --region us-central1 \
  --allow-unauthenticated \
  --set-env-vars GEMINI_API_KEY=your_key_here

📁 Project Structure

ASCA/
├── backend/
│   ├── agents/
│   │   ├── assignment_analyzer.py
│   │   ├── schedule_optimizer.py
│   │   └── wellness_monitor.py
│   ├── main.py
│   ├── test_agents.py
│   ├── requirements.txt
│   └── Dockerfile
├── index.html
├── api-client.js
└── README.md

🎥 Demo

[Link to demo video]

📝 License

MIT License

🏆 Hackathon

Built for [Hackathon Name] - Multi-Agent AI Challenge