Ayan User

I’m a systems and data-focused developer with hands-on experience building high-performance C++ applications and SQL-driven pipelines. I’m comfortable jumping into performance-critical, network-aware code while keeping data flows reliable and measurable.

In my Deep Packet Inspection (DPI) Engine project, I built a multi-threaded packet analysis engine using POSIX Threads and PCAP, with dynamic thread-pool load balancing. I delivered sub-2 ms per-packet latency and built a 5-tuple session reconstruction pipeline for concurrent classification with stateful tracking.

On the data and modeling side, I engineered a factor-based Quantitative Equity Return-Prediction Model using Python, SQL, and XGBoost. I achieved 63.4% directional accuracy on out-of-sample test data and validated performance through a complete backtest with portfolio P&L, Sharpe Ratio, and maximum drawdown.

I also created “Adversarial-PCG-SLM: Spatial Reasoning Framework,” where I fine-tuned via Hugging Face and used SQL-backed logging to stress-test an SLM’s spatial reasoning limits in a procedurally generated 2D maze. Through my Software Engineering Virtual Experience at JP Morgan Chase & Co. (Forage) in 2024, I built a real-time stock price data feed and delivered end-to-end dashboard visualisation using Agile practices and Git-based version control.