MindBoost

MindBoost is an intelligent academic companion platform built with a microservices architecture.
It helps students and educators analyze study materials, generate knowledge graphs, create quizzes, track learning progress, and monitor academic well-being.

The system integrates AI-powered document parsing, knowledge extraction, quiz generation, burnout prediction, and interest tracking, all accessible via a modern React frontend orchestrated with Docker Compose.

---

Table of Contents

• Overview
• Architecture
• Microservices Included
• Technologies Used
• ML Model: Burnout Predictor
• Running the System
• Running Microservices Individually
• Test Users
• Initial Setup
• Versioning & Contribution

---

Overview

MindBoost combines data processing, AI inference, and interactive dashboards to assist academic users in managing knowledge and maintaining a healthy balance between learning and well-being.

Core functionalities include:

• Parsing and analyzing uploaded study materials.
• Generating knowledge graphs to visualize concepts.
• Creating and scoring AI-generated quizzes.
• Monitoring user interest and burnout risk through machine learning.
• Providing real-time analytics via Streamlit dashboards.

---

Architecture

All services are containerized and orchestrated via Docker Compose.
Below is a simplified system diagram illustrating the service interactions:

graph TD
A[Frontend] --> B[Backend API Gateway]

C[PDF Parser Microservice] --> E[Bedrock Client Microservice]
C --> D
B --> D[Knowledge Graph Microservice]
B --> E
D --> J
B --> F[Quiz Burnout Gateway]
F --> G[Quiz Score Microservice]
G --> L[(MongoDB)]
G --> K[Interest Monitor Streamlit]
K --> O[(MongoDB)]
F --> H[ML Model Burnout]
H --> I[Trained Model]
E --> J[Bedrock Monitor Streamlit]

Loading

Flowchart

flowchart TD
    subgraph User Interaction
        A[Login ]
        C[PDF Upload]
        M[Burnout Detection Button]
    end

    subgraph Processing
        D[PDF Parser Microservice]
        E[Extracted Text]
        F[Knowledge Graph Microservice]
        H[Bedrock Client Microservice]
        O[ML Model Burnout Microservice]
        Q[Bedrock Monitor Microservice]
    end

    subgraph Storage
        K[(MongoDB)]
    end

    subgraph Output
        G[Knowledge Graph Visualization]
        I[Quiz Generation]
        J[Quiz Score Microservice]
        L[User Takes Quiz]
        P[Burnout Risk Result]
        R[Usage & Cost Analytics]
    end

    A --> B[Backend]
    A --> C
    C --> D
    D --> E
    E --> F
    F --> G
    E --> H
    H --> I
    I --> J
    J --> K
    I --> L
    L --> J
    M --> N[Quiz Burnout Gateway]
    N --> K
    N --> O
    O --> P
    H --> Q
    Q --> R

Loading

---

Microservices Included

Service	Technology	Description	Port
backend	FastAPI	Main API Gateway, routes all requests	8000
pdf-parser-microservice	FastAPI	Parses PDF files and extracts text	8001
bedrock-client-microservice	FastAPI + Boto3	Connects to AWS Bedrock for LLM tasks	8002
knowledge-graph-microservice	FastAPI + NetworkX	Generates knowledge graphs from parsed text	8003
ml-model-burnout	Flask	Predicts burnout probability using ML model	8004
quiz-score-microservice	Flask	Computes quiz results and stores them in MongoDB	8010
quiz-burnout-gateway	Flask	Aggregates quiz data and calls burnout model	8011
interest-monitor-microservice	Flask + Streamlit	Displays user topic interests and analytics	8502
mongodb	MongoDB	Stores quiz results, user data, and interests	27017
bedrock-monitor	Streamlit	Dashboard for Bedrock Client metrics	8501
frontend	React + MUI	Web interface for interacting with MindBoost	3000

---

ML Model: Burnout Predictor

The ml-model-burnout microservice predicts a user’s likelihood of academic burnout using a trained model.

Model Details

• Model: model.pkl (pre-trained using joblib)
• Input: Time series of scores (e.g., stress, wellbeing)
• Output: Probability of being close to burnout (0–1)

Input Example

{
  "user_id": "u1",
  "series": [
    {"date": "2025-01-05", "score": 82},
    {"date": "2025-01-12", "score": 78},
    {"date": "2025-01-20", "score": 74}
  ]
}

Output Example

{
  "user_id": "u1",
  "prob_close_to_burnout": 0.78,
  "features": { ... }
}

---

Running the System

You can run the entire MindBoost platform using Docker Compose.

# Build and start all services
docker-compose up --build

After all services start:

• Frontend: http://localhost:3000
• Bedrock Monitor: http://localhost:8501
• Interest Monitor: http://localhost:8502
• API Gateway: Accessible at http://localhost:8000

---

Running Microservices Individually

Each service can be started separately for local development:

Service	Directory	Command
Backend (Gateway)	src/backend	uvicorn src.main:app --reload
PDF Parser	src/pdf-parser-microservice	uvicorn src.main:app --reload
Bedrock Client	src/bedrock-client-microservice	uvicorn src.main:app --reload
Knowledge Graph	src/knowledge-graph-microservice	uvicorn src.main:app --reload
ML Model Burnout	src/ml_model_burnout	flask --app predict_service run --host=0.0.0.0 --port=8004
Quiz Score	quiz-score-microservice	flask --app main run --host=0.0.0.0 --port=8010
Quiz Burnout Gateway	quiz-burnout-gateway	flask --app main run --host=0.0.0.0 --port=8011
Bedrock Monitor	bedrock-monitor	streamlit run app.py
Interest Monitor	interest-monitor-microservice	streamlit run app.py

---

Test Users

Use these credentials during development and testing:

Email	Password
[email protected]	password123
[email protected]	letmein456
[email protected]	qwerty789

---

Initial Setup

Before logging in for the first time, populate the local SQLite and MongoDB databases.

# Install dependencies
pip install -r misc/requirements.txt

# Run the database initialization script
python misc/init_db.py

---

Versioning and Contribution

Versioning

• Python: 3.9 / 3.10
• Node: 18+
• Dependencies managed via requirements.txt and package.json.

Contributing

Pull requests are welcome!
Please open an issue to discuss proposed changes before submitting.

---

Summary

MindBoost provides:

• Modular AI-powered learning analysis
• Seamless REST-based microservice communication
• Real-time dashboards for insightful monitoring
• A robust and scalable Dockerized environment

---

Demo