Projects
A collection of my work showcasing expertise in high-frequency trading systems, full-stack development, and scalable application architecture.
DefenseAI
Defense-grade AI decision support system that demonstrates human-in-the-loop design, safety-first AI orchestration, and production-grade engineering practices. It is designed to generate structured intelligence briefs from raw source data while ensuring that no AI output is actionable without human approval.
SynapseStrike
SynapseStrike AI trading system that lets you run multiple AI models to trade automatically. Configure strategies through a web interface, monitor performance in real-time, and let AI agents compete to find the best trading approach.
PPOAlgo - Momentum Algorithm Using RTX 3090
Modern trading strategies live or die by their parameters. Buy too early, sell too late, size incorrectly—and even a good idea collapses. The real challenge isn’t coming up with strategies; it’s finding the optimal parameters behind them. That’s exactly why I built PPOAlgo.
Invest-Soft – Algorithmic Stock Trading Platform
A comprehensive stock trading platform and algorithmic trading dashboard featuring real-time market data, automated trading algorithms, portfolio management, and advanced analytics.
Evolutionary Strategy Research
A research framework for evolving trading strategies using genetic programming and evolutionary algorithms. This project provides tools for automatically discovering strategy building blocks, running reproducible experiments, and evaluating candidate strategies through backtesting.
Quant Strategy Simulator: 155% return in 1 year
Based on my previous algo trading application i developed Stock Optizmier which scans hundreds of stocks and test them based on Vwap extended algo, it runs over 1 mil operations per stock and it backtest with a differetn data. Based on sorting i found a stock which return 155% in return in 1 year based on last year backtesting
C++ Low-Latency Trading Core
Goal: Prove low-latency C++ ability for infra-level work. Stack: C++17, Boost.Asio, ZeroMQ, pybind11 (for Python bindings) Key Features: Implement order book and matching engine REST + WebSocket API layer Multi-threaded architecture Nanosecond-level latency tracking Optional replay mode for historical data
Market Data Pipeline (ETL System)
Goal: Show data engineering ability and understanding of trading data flow. Stack: Python, Airflow or Prefect, Kafka or Redis Streams, PostgreSQL, Docker Key Features: Stream data from public APIs (Polygon.io, Binance, Alpaca) Normalize to uniform schema Store in PostgreSQL or Parquet files Schedule ETL jobs via Airflow/Prefect Simple web dashboard for monitoring pipeline health
Historical Market Data API
WideSurf API Platform is a production-grade stock market data service designed as a scalable alternative to platforms like www.alpaca.markets. Built with modern cloud-native architecture and deployed on Kubernetes (K3s), it delivers high-performance financial data APIs alongside a polished, user-friendly dashboard.
Trading Metrics & Monitoring Dashboard
Goal: Show observability and DevOps maturity. Stack: Python, Grafana or Streamlit, Prometheus Key Features: Collect latency, PnL, win/loss, and system metrics Visualize in real-time charts Create alert rules for abnormal latency or drawdown Integrate with the trading engine
AI Architect
I designed and built a local-first AI decision engine that mirrors how an experienced trader thinks—step by step—while keeping full control over data, cost, and risk. At the core is a 6-stage reasoning pipeline I call the Trading Brain, where each stage adds context, validation, and accountability before a decision is made.
TaskFlow
A production-ready multi-tenant task management SaaS platform built with FastAPI and Next.js. Features include data pipeline engineering with Celery workers, real-time WebSocket updates, comprehensive analytics, and organization-based data isolation.
LocalAI - GPT-OSS-20B
LocalAI - GPT-OSS-20B which works on my rtx 3090 GPU and works without accessing to the internet.
FinGPT GUI
An advanced, containerized AI agent for real-time stock analysis, multi-model forecasting, and high-alpha opportunity hunting.