Available for consulting

Huang
Tang

End-to-end DS/ML/GenAI product developer who brings innovative and practical solutions to realistic business challenges — from domain modeling to deployed systems.

View Projects Get In Touch Resume ↗
8+ Years Experience
6 Domains
20+ Products Shipped
01

About

I'm a full-stack data scientist and ML engineer with a systems-first mindset. My work begins with domain knowledge — understanding the actual physics, economics, or operational logic of a problem — and builds up through mathematical modeling to production-ready software.

I specialize in bridging the gap between research and deployment: turning rigorous mathematical models into scalable GenAI pipelines, decision systems, and ML products that solve real business problems.

Whether it's a clinical decision support tool, an aviation safety classifier, or a financial forecasting engine, I bring the same disciplined approach: first principles, then code.

Domain coverage

Healthcare & Clinical Decision Support
Aviation Safety & Operations
Financial Modeling & Risk
Decision Intelligence
Network Operations
Enterprise GenAI Products
02

Experience

2026 — Present
Lead GenAI Engineer
[Fannie Mae]
Lead end-to-end development of GenAI products — from problem framing and mathematical modeling to production deployment. Designed multi-agent reasoning systems and RAG pipelines for enterprise clients across healthcare and finance.
2025 — 2026
Principal Artificial Intelligence Architect
[REI Systems]
Built predictive models and decision-support tools across aviation safety and network operations domains. Architected scalable ML pipelines using Python and TypeScript, reducing model deployment cycles by 40%.
2012 — 2025
Principal Artificial Intelligence Engineer
[The MITRE Corporation]
Developed statistical models and NLP systems for business intelligence. Led cross-functional teams in translating domain expertise into quantitative solutions. Delivered 10+ production ML systems.
03

Skills

Languages
Python TypeScript SQL R Bash
ML / AI
PyTorch scikit-learn HuggingFace LangChain LlamaIndex Transformers
GenAI
RAG Systems Multi-agent Fine-tuning Prompt Eng. Evals
Data Engineering
Spark dbt Airflow Kafka PostgreSQL Redis
Cloud / MLOps
AWS GCP Docker Kubernetes MLflow FastAPI
Methods
Bayesian Inference Time Series Causal ML Optimization NLP
04

Projects

Healthcare
Clinical Decision Support Engine
End-to-end RAG-powered diagnostic assistant that synthesizes patient records, clinical guidelines, and real-time lab data to surface evidence-based treatment recommendations.
Python LlamaIndex GPT-4 FastAPI PostgreSQL
Aviation
Flight Safety Anomaly Detector
Temporal convolutional network trained on FOQA data to identify precursor patterns of safety events up to 90 seconds before occurrence, enabling proactive crew alerts.
PyTorch TCN Spark Airflow AWS
Finance
Probabilistic Risk Forecasting System
Bayesian hierarchical model for credit risk forecasting across portfolio segments, with uncertainty quantification and scenario simulation for regulatory stress testing.
Python PyMC TypeScript React GCP
GenAI
Multi-agent Enterprise Copilot
Orchestrated multi-agent system with specialized sub-agents for data retrieval, code execution, and domain reasoning — deployed as an internal knowledge and productivity tool.
LangGraph Claude TypeScript Next.js Redis
05

Blog

2025-03-12 Designing RAG systems for clinical workflows: lessons from production GenAI 2025-01-28 Why causal graphs matter before you touch your data Methodology 2024-11-05 Uncertainty quantification in ML: from Bayesian posteriors to prediction intervals ML Theory 2024-09-17 Multi-agent orchestration patterns for enterprise GenAI Architecture 2024-07-02 Domain-first modeling: why I start with physics, not data Philosophy
06

Contact

Let's build
something rigorous.

I'm open to consulting engagements, full-time roles, and research collaborations — especially in healthcare, aviation, finance, or any domain where the problem is genuinely hard.

Preference for problems that start from first principles.

Email huangtang@gmail.com GitHub github.com/huangtang-lst LinkedIn linkedin.com/in/huangtang