Hi, I'm

Mina Lee

AI Engineer · Agentic solutions & LLM systems

6+ years in software engineering, 4+ years delivering LLM-powered applications and building greenfield AI infrastructure. I bridge research and production: turning model capabilities into reliable, scalable systems. Currently at Microsoft on the Azure Core AI team. MS in Computer Science from NYU.

Mina Lee

Skills

Languages

Python Java Go C++ TypeScript SQL C#

Frameworks & Libraries

PySpark Hugging Face NeMo Curator React Flask SpringBoot gRPC TensorFlow

AI & LLM

Agentic Systems Multi-agent Workflows LLM Fine-tuning Prompt Engineering RAG LLM-as-a-Judge OpenAI API

Cloud & Infra

Azure AWS GCP Kubernetes Docker MySQL MongoDB Redis

Projects

Agentic Mainframe Modernization

End-to-end agentic system for migrating legacy COBOL to Java using LLM models, productionized for enterprise private preview.

LLM models Prompt Engineering Azure Applied AI

Multilingual LLM Data Pipeline

Privacy-safe multilingual data preparation pipeline for LLM training, processing 150M+ documents (90B tokens).

NeMo Curator Docker Data cleaning Deduplication

LLM Evaluation & Benchmarking System

In-house platform to benchmark LLM checkpoints across configurations, with a conversational AI agent for orchestration and researcher dashboards.

Python Kubernetes Docker React SQL Hugging Face EleutherAI

Intelligent Anomaly Detection

Monitored Azure SQL service metrics and refined precision-recall of ML-based detections in collaboration with data scientists.

Python Machine Learning

Alexa NLP Deployment Sandbox

Testing platform for NLP model deployment in the Alexa ML pipeline — automated model release via A/B testing, improving branch coverage by 20%.

Java NLP A/B Testing AWS

E-Scooter Demand Forecasting Research

Forecasting model for the impact of e-scooter sharing deployment in Manhattan.

Published paper →
Informatics analysis Data Science

Get in Touch

I'm always open to interesting conversations and new opportunities.