Skip to content

usk2003/StockTraQ-v2

Repository files navigation

Stock TraQ

AI-powered IPO Analysis & Archive Engine

Stock TraQ is a professional-grade platform designed to provide deep insights into IPO performance. It leverages advanced machine learning models to predict listing gains and audit financial health, while maintaining a comprehensive archive of historical IPOs.


Why StockTraQ is Unique (vs. General AI Chatbots & Tools)

While generative AI models (like ChatGPT) are suited for text synthesis and general descriptions, StockTraQ addresses critical shortcomings in financial predictive modeling:

  1. Analytical Regression vs. Generative Text: AI chatbots rely on language correlations prone to hallucinating numerical estimates. StockTraQ executes rigorous local mathematical regression using local pre-trained weights for exact deterministic outputs over verified data.
  2. Interactive Analytic Dashboard: Standard AI tools cannot render unblocked sliders, continuous visual charts, or calculate threshold unified aggregates efficiently over layered datasets.
  3. Ensemble Modeling over Approximations: Common calculators use static averages; StockTraQ combines Random Forest, Gradient Boosting, and Linear Regression ensembles concurrently safeguarding outputs from statistical outliers.

Key Features

  • Unified IPO Rating: A hybrid 1-10 score generated by 5 specialized ML models.
  • Archive Explorer: Interactive search and audit for historical listings.
  • AI Performance Audit: Real-time backtesting comparing AI predictions vs. actual historical gains.
  • Neural Engine: Combines Random Forest, Gradient Boosting, and Linear Regression for high-precision forecasting.
  • Premium Interface: A glassmorphic, responsive UI built for modern financial intelligence terminals.

Technology Stack

  • Frontend: React 18 (Vite), Tailwind CSS, Lucide Icons, Framer Motion.
  • Intelligence Backend: FastAPI (Python 3.10+), Pydantic, Scikit-learn.
  • Management Backend: Node.js (Express), JSON Web Tokens (JWT) for Admin/User Auth.
  • Database: MongoDB (Archival data storage).

AI Prediction Models

  1. Listing Gain Predictor: Forecasts opening day performance based on subscription tiers (QIB, NII, Retail).
  2. Financial Strength Audit: Scores company fundamentals using Revenue, PAT, ROE, and ROCE metrics.
  3. Valuation Impact: Analyzes pricing efficiency relative to P/E ratios and sector trends.
  4. Demand Tier Classification: Categorizes market interest from 'Low' to 'Blockbuster'.
  5. Long-Term Projection: Estimates performance trends over a 6-12 month horizon.

Getting Started

Prerequisites

  • Node.js (v16+)
  • Python (v3.10+)
  • MongoDB (Local or via Docker)
  • Docker & Docker-Compose (Recommended for easiest setup)

🟢 Option A: Run with Docker (Fastest Method)

If you have Docker installed, simply run the ecosystem with:

docker-compose up --build

Note: This automatically orchestrates the Node.js API, FastAPI ML engine, and the Client bundle.


🔵 Option B: Manual Setup Alternative

1. Database Setup

Ensure MongoDB is running and contains the setup tables for ongoing_ipos, closed_ipos, and master_db.

2. Intelligence Backend Setup (FastAPI)

cd backend
python -m venv .venv
source .venv/bin/activate  # or .venv\Scripts\activate on Windows
pip install -r requirements.txt
python main.py

3. Management Backend Setup (Node.js)

cd node-backend
npm install
node server.js

4. Frontend Setup (React)

cd frontend
npm install
npm run dev

5. Running the Complete App

To launch all services concurrently, use the provided batch script located in the project root:

.\run.bat

Methodology

Stock TraQ utilizes a time-based data split to simulate real-world forecasting:

  • Training Set: Historical IPO data.
  • Validation Set: High-volatility listings.
  • Ensemble Weights: 40% Random Forest, 40% Gradient Boosting, 20% Linear Regression for fallback execution.

Disclaimer

Investment Disclaimer: Stock TraQ provides AI-based predictions for educational and research purposes only. IPO investments carry significant market risk. Predictions and ratings should not be considered financial advice. Always consult with a certified financial advisor before making investment decisions.


Built for Modern Investors.

About

Stock TraQ is a professional-grade platform designed to provide deep insights into IPO performance. It leverages advanced machine learning models to predict listing gains and audit financial health, while maintaining a comprehensive archive of historical IPOs.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors