I build end-to-end quantitative research systems: raw data ingestion, factor modeling, backtesting, and risk analysis.
My main track is quant.
My secondary track is cybersecurity, with an interest in secure and reliable systems.
I care about:
- extracting signal from noisy markets
- reproducible research pipelines
- modeling real-world constraints instead of toy results
|
Fully reproducible replication of the 12–1 month cross-sectional momentum strategy in the S&P 500 (2005–2024), with survivorship-bias-free data and realistic transaction costs. |
Modular algorithmic trading framework for live trading (Alpaca) and historical backtesting, with automated performance reporting, SPY benchmarking, and reproducible trade logs. |
|
Real-time network attack detection platform with a 5-rule engine and a 3-model ML ensemble (Isolation Forest, Autoencoder, K-Means) on a live React dashboard. |
Empirical analysis of limit order book dynamics: order flow imbalance, return predictability, signal decay, and market microstructure effects using Kraken spot data. |
[2026-06] building netguard — full-stack IDS with ML ensemble detection
[2026-05] limit order book microstructure analysis on Kraken spot data
[2026-04] published momentum-factor-study with reproducible pipeline and LaTeX paper
[2026-03] improving turnover-aware backtesting assumptions
languages python | typescript | r | lua
libraries pandas | numpy | scikit-learn | statsmodels | fastapi
tools docker | postgresql | redis | alpaca api | github actions
methods time-series | econometrics | ml ensemble | backtesting | simulation
- published quantitative research with full reproducibility
- built a full-stack IDS with multi-model ML detection
- interested in alpha research, market microstructure, and robust systems
- building at the intersection of quant finance and cybersecurity infrastructure
- Portfolio: https://dshan12.github.io
- GitHub: https://github.com/dshan12
- LinkedIn: https://www.linkedin.com/in/darshansathishkumar/
- Email: darshansathishkumar@gmail.com
markets are adaptive systems. edge comes from structure.


