I am a Software Engineer specializing in Full Stack Development and Machine Learning. My experience lies in building high-performance systems and data-intensive applications within trading environments, ranging from real-time financial dashboards to analysing huge datasets to squeeze insights.
Currently exploring Generative AI and Agentic Systems to develop LLM-based agents for decision-making and workflow automation.
Seeking opportunities to leverage my skills and experience in Fullstack development, Python, and Rust to build scalable, data-intensive products.Open to connecting!
- Frontend: React.js, Vue.js, JavaScript (ES6+), Tailwind CSS, Grafana
- Backend: Node.js, Python (FastAPI/Flask), Rust, Go, REST & FIX Protocols
- Data & ML: PyTorch, Scikit-learn, Pandas, Polars, PostgreSQL, DynamoDB
- Tools: Docker, Git, Jenkins, Vector.dev, Linux Systems
- Real-Time Dashboards: Architected and deployed end-to-end data visualization platforms using Vector.dev, Python and Grafana to track markets in real-time.
- High-Performance Backends: Developed multithreaded protocol parsers and backtesting utilities in Rust and Python (Polars) to process massive datasets where python reached its limits.
- Data Pipelines: Built real-time dashboards, to analyze market efficiency across thousands of instruments simultaneously.
- Automation: Developed a command-line interface for real-time configuration generation, streamlining the deployment of complex market parameters.
- Full Stack Development: Built a centralized lifecycle management platform using React and Flask, integrating external APIs to monitor software security and status across distributed systems.
- Client-Side Engineering: Redesigned monitoring agents from Python to Go, creating lightweight, compiled binaries for zero-downtime performance.
- Chrome Extensions: Developed specialized internal tools , a browser extension for inventory reconciliation.
A classified posting application featuring notification systems and GeoJSON integration for location-based search and display.
An automated trading engine integrated with the Oanda API, focusing on robust execution logic and real-time risk management.
Achieved a rank of 181 out of 812 in a competitive ML challenge. Engineered LSTM and Auto ARIMA models to predict multi-stock volatility and closing prices.
A library management system featuring a recommendation engine, showcasing the integration of ML-driven insights into a standard full-stack application.
- LinkedIn: Piyush Soni
- Twitter: @ViewsOfPiyush
- Portfolio: Visit My Repositories
- Email: me.piyush.360@gmail.com
"Building the bridge between data-driven insights and seamless user experiences."