Quantitative research infrastructure for financial data ingestion, analysis, and visualization.
QuantCore AI is a full-stack research platform designed for structured quantitative workflows.
It provides authenticated project environments where users can ingest market data, validate datasets, compute return statistics, and visualize price dynamics through a modular backend and interactive analysis interface.
The platform focuses on reproducible data pipelines, clear system architecture, and research-grade analysis tooling rather than single-strategy implementations.
QuantCore AI provides the core components needed for exploratory financial research:
- Secure user authentication (JWT-based)
- Project-scoped research environments
- Market data ingestion (CSV – Yahoo/NSE formats)
- Automatic dataset validation and cleaning
- Dataset storage and version tracking
- Exploratory data analysis (EDA)
- Daily return computation
- Annualized return estimation
- Volatility estimation
- Interactive candlestick price visualization
- Returns time-series visualization
- Exportable cleaned datasets
These components form the foundation of a reproducible quantitative research pipeline.
QuantCore AI is designed around three principles common in quantitative research environments.
Reproducibility Data ingestion, validation, and analysis workflows are structured so results can be replicated across projects and datasets.
Data Integrity Financial time-series datasets are automatically validated and cleaned to avoid silent errors that distort statistical analysis.
Separation of Research Layers The platform separates data ingestion, analysis logic, and visualization layers, enabling modular experimentation and easier extension into systematic strategy research.
Frontend (Streamlit) ↓ REST API (FastAPI) ↓ Research Modules ├── Authentication (JWT) ├── Project Management ├── Dataset Ingestion ├── Data Validation & Cleaning └── Statistical Analysis ↓ Database (SQLite – MVP)
Each user operates within isolated project environments where datasets and analyses remain scoped to that workspace.
- FastAPI
- SQLAlchemy
- Pydantic
- JWT Authentication
- SQLite (MVP database)
- Pandas
- NumPy
- Streamlit
- Plotly (interactive visualizations)
- OHLC financial time-series
- Yahoo Finance / NSE style CSV datasets
Current analytical tools include:
- raw dataset preview
- cleaned dataset generation
- daily return computation
- annualized return estimation
- volatility estimation
- candlestick price charts
- returns time-series visualization
- downloadable cleaned datasets
These provide a structured exploratory layer before more advanced modeling or strategy research.
- User logs in
- Creates a research project
- Uploads market CSV data
- Dataset is automatically validated and cleaned
- Platform computes returns and volatility statistics
- User explores price behavior through charts and summaries
- Cleaned datasets can be exported for downstream modeling
Clone the repository git clone https://github.com/Aryan-core/quantcore-ai.git cd quantcore-ai
Create virtual environment python -m venv venv source venv/bin/activate
Install dependencies pip install -r requirements.txt
Start backend uvicorn backend.main:app –reload –port 8000
API documentation available at: http://127.0.0.1:8000/docs
Start frontend streamlit run frontend/app.py
Application runs at: http://localhost:8501
MVP v0.2
The current version focuses on:
- correct data handling
- modular system architecture
- reproducible research workflows
This stage intentionally avoids strategy optimization or performance claims.
Planned extensions include:
- strategy backtesting engine
- equity curve and PnL analysis
- additional risk metrics (Sharpe ratio, drawdown)
- multi-dataset research projects
- cloud deployment
- research notebook integration
Aryan Khan
Drexel University
Founder — QuantCore AI
This project is intended for research and educational purposes only.
It does not constitute financial or investment advice.