I'm a computational engineer building and deploying full-stack ML and LLM-based systems. I hold a PhD from the University of Texas at Austin, where I developed a strong foundation in numerical methods, stochastic modeling, and scientific computing. My work spans feature engineering, multi-objective optimization, regime detection, stochastic modeling, RAG, and multimodal document intelligence.
I've shipped production trading systems, LLM-backed research agents, market intelligence tools, and cloud-native analytics applications. I care about the full lifecycle from data pipelines and model development through deployment, evaluation, and monitoring.
I built production ML systems and trading infrastructure for quantitative strategies — volatility arbitrage intraday options trading on AWS Lambda, Chronos embeddings with LoRA tuning and AutoGluon-backed interday options trading on Modal, and NSGA-II-backed sector ETF portfolio construction with regime-adaptive rebalancing on AWS Lambda, all wired together with GitHub CI, Docker, EventBridge scheduling, React dashboards, and DynamoDB.
On the LLM side, I engineered a Groq-powered FX and macro RSS feed-driven intraday trading agent on AWS Lambda, built a FAISS RAG-powered Groq agent for swing trading of commodity stocks on Modal, and deployed a news-backed LLM-triggered rebalance workflow for a Quantum Computing and AI thematic index. I also designed a query-based trade signal mining app over backtested options data using an LLM-guided workflow, and engineered a Groq vision model and two-pass LLM-enabled PDF Q&A app for market research.
Alongside these systems I experimented with single-pass vs. multi-pass LLM prompt architectures, retrieval configurations and embedding model choice for the RAG pipeline, Pareto front selection rules for bi-objective optimization, solver benchmarking for stochastic volatility model calibration, and A/B tests comparing SHAP vs. native feature importance and LoRA fine-tuning vs. frozen Chronos embeddings.
I joined during the early integration phase of a robotic surgical navigation platform combining imaging and robotic actuation. My work focused on implementing TCP/UDP communication protocols for robotic arm motion control, contributing to end-to-end packaging with Poetry, and supporting deployment through Azure tooling. The team delivered a successful live system demonstration used in Series A fundraising.
I adapted physics-informed modeling originally developed for fusion experiment data to financial time series. The resulting production pipeline combined sparse regression for PDE construction, signal generation, and automated execution. I also implemented a CAGR-maximizing Bayesian TPE optimizer for swing-trading thresholds and deployed the system on AWS Lambda.
ML & Optimization
Model development, feature engineering, and optimization across classical ML, probabilistic models, and evolutionary search.
Cloud & MLOps
Serverless-first production systems with CI/CD, infrastructure-as-code, containers, and model deployment workflows.
Programming
Python-first engineering with scientific computing, fast data processing, and frontend work for internal analytics tools.
APIs & Web
Production APIs and app backends connecting market data, broker integrations, feeds, and interactive analytics interfaces.
NLP & LLMs
RAG systems and LLM tooling with observability, vector retrieval, and evaluation for production use.
Beyond the code
