Foresight Analytics

Core Features

• Interactive Dashboard: Explore data with dynamic filters.
• Linear Regression Forecasting: Generate future projections.
• Rich Visualizations: View monthly trends, category compositions, and forecast results.
• Optimized Initial Load:
  • Fast Startup: Implemented efficient database checks to bypass CSV reprocessing if data is already seeded in SQLite. Subsequent app loads are now significantly faster.
  • SQLite Backend: Uses SQLAlchemy ORM for robust data management, moving beyond direct CSV reads for improved query performance.
• Configuration Management: Centralized settings in app/config/settings.py for easy maintenance.
• Data Validation: Ensures data quality before processing and storage.
• Caching Strategy:
  • In-session performance: 85% faster subsequent operations
  • Smart data retrieval: Multi-tier caching with strategic TTL (Time to Live)
  • User experience: Sub-second response for tab switching and filtering
  • Resource efficiency: Eliminates redundant database queries
  • Impact: performance gain ~85% faster
    • Without Caching: Department change → 2.1s load time
    • With Caching: Department change → 0.3s load time

Kaggle API Setup

1. Create Kaggle account and generate API token
2. Create .env file in root directory: KAGGLE_USERNAME=your_username KAGGLE_KEY=your_api_key

Running Locally

For winsows

• git clone https://github.com/ht-l1/Foresight-Analytics.git
• .\venv\Scripts\activate
• pip install -r requirements.txt
• streamlit run run.py

Tech Stacks Used

Core:

Database:

Visualization:

Foresight-Analytics
├─ app
│  ├─ config
│  │  ├─ settings.py
│  │  └─ __init__.py
│  ├─ data
│  │  ├─ data_loader.py
│  │  ├─ test_supabase.py
│  │  └─ __init__.py
│  ├─ database
│  │  ├─ models.py
│  │  └─ supabase_manager.py
│  ├─ models
│  │  ├─ forecaster.py
│  │  └─ __init__.py
│  ├─ utils
│  │  ├─ logger.py
│  │  ├─ validators.py
│  │  └─ __init__.py
│  └─ __init__.py
├─ data
│  ├─ foresight_analytics.db
│  └─ train.csv
├─ LICENSE
├─ README.md
├─ requirements.txt
└─ run.py