Ethan F. Cooley — ML Engineer

00 / About

The
Person

Ethan F. Cooley35.2271° N

Ethan Cooley is a Senior Machine Learning Engineer based in Charlotte, NC, building AI systems that improve how clinicians make decisions. He studied engineering and computer science at Wake Forest University and is currently pursuing a Master's in AI at Duke University's Pratt School of Engineering.

Outside of work he can be found running ultras in the mountains, competing in triathlons, skiing backcountry, or shredding trails on his mountain bike. He's a brother, a son, and a friend — the people in his life matter more than any finish line or job title.

Charlotte, NC Wake Forest University · Duke MEng AI (in progress) Ultramarathons · Backcountry Skiing · MTB

01 / Experience

Professional
History

Stanson Health

A division of Premier Inc. · Charlotte, NC

Senior Machine Learning Engineer July 2025 — Present

Optimize end-to-end LLM inference pipeline through kernel caching and batched sentence embedding strategies, reducing latency and cutting compute costs.
Architect and continuously expand a comprehensive functional test suite for LLM models, improving end-to-end coverage and ensuring reliability across model updates and pipeline changes.
Lead design, development, and maintenance of the data engine for the data science team, leveraging AWS OpenSearch as a centralized tool for efficient data interaction and retrieval.
Architect and implement RESTful API endpoints connecting front-end applications with ML models deployed in AWS SageMaker, enabling rapid access to model predictions and enhancing user experience.
Collaborate cross-functionally with clinical product teams and software engineers to define project requirements and deliver solutions aligned with business objectives.

Machine Learning Engineer September 2023 — July 2025

Designed and implemented an internal annotation tool enabling healthcare providers to efficiently label cases, directly improving training data quality for production ML models.
Established team-wide coding standards by introducing Black and Flake8 across all repositories, fostering a culture of maintainable, consistent code.
Built monitoring and logging solutions to track ECR pipeline and SageMaker model performance, leveraging AWS CloudWatch for real-time insights and proactive issue resolution.
Authored comprehensive documentation for data engine architecture and API endpoints, and led training sessions to onboard team members to new tools and workflows.

Data Scientist June 2021 — September 2023

Built active learning infrastructure that queried uncertain cases from production and surfaced them for annotation, continuously improving model accuracy over time.
Employed directed acyclic graphs (DAGs) to streamline and automate annotation jobs, significantly enhancing efficiency of case annotation for ML model training.
Utilized NLP techniques to assist clinicians in making medical predictions derived from electronic health record data, directly improving patient care outcomes.

Data Science Intern January 2021 — May 2021

Collaborated with senior data scientists to iterate on and improve production ML models, gaining hands-on experience with the full model development lifecycle.
Developed and enhanced an internal software app for testing and comparing model predictions, strengthening team understanding of model strengths and weaknesses.

02 / Skills

Technical
Stack

Filled tags are what I use daily. Outlined tags I'm also comfortable with.

Machine Learning

Python LLMs RAG NLP Sentence Embeddings Active Learning PyTorch Hugging Face TensorFlow Scikit-Learn Transfer Learning CNNs EfficientNet

Cloud & Infrastructure

AWS SageMaker AWS OpenSearch AWS ECR AWS S3 AWS CloudWatch Google Cloud Firebase Vercel

Software Engineering

RESTful APIs OAuth 2.0 React Node.js JavaScript HTML / CSS Flask FastAPI Git

Data & MLOps

Data Pipelines DAGs Model Monitoring NumPy Pandas SQL A/B Testing

03 / Education

Academic
Background

Duke University

Master of Engineering — Artificial Intelligence for Product Innovation

Jul 2025 — May 2027 In Progress

'27

Wake Forest University

B.S. in Engineering — Cum Laude

May 2021 · Winston-Salem, NC

Won best senior capstone project overall — prototyping, presentation, pro-humanitate design, and final delivery. View project →

'21

University of Sydney

Study Abroad — School of Engineering & Computer Science

Fall 2019 · Sydney, Australia

'19

04 / Projects

Featured
Work

Trace

Personal Project

2025 · React, Node.js, Vercel

Trace is a full-stack activity intelligence dashboard that connects to the Strava API and aggregates training history across runs, rides, and swims into a clean, filterable interface with summary statistics.
Utilizes a secure OAuth 2.0 token exchange flow via a Node.js serverless backend, with automatic token refresh handling so sessions stay alive without re-authentication.
Renders GPS route traces as pure canvas visualizations — no base maps, just the raw polyline — with layered glow effects and an animated card-to-fullscreen transition driven by live screen coordinates.
Surfaces sport-specific performance metrics per activity including pace, power, normalized power, heart rate, elevation gain, cadence, and suffer score.

View live ↗

Live canvas · click to replayGPS trace

From Elite to Everyone

Personal Writing · Data Analysis

Medium · September 2025 · Python, Pandas, Data Visualization

Analyzed 7.4 million ultramarathon results from 1.6 million athletes, spanning 1798–2022, to understand how the sport's participation and performance have evolved over time.
Used pandas for feature engineering and time-series analysis, uncovering a steady rise in finish times since the 1970s — driven not by declining performance, but by a surge of first-time, older, and more geographically diverse participants.
Visualized total finishers, age distribution, and country diversity across seven decades to make the case that slower average times reflect a sport becoming more inclusive rather than less competitive.

0M+Race results analyzed

0Years of data (1798–2022)

0+Countries by 2020

Read on Medium ↗ View code ↗

Trail runner descending a ridge above Chamonix, glaciers and clouds in the background

Read on Medium ↗Photo: Brian Metzler / Unsplash

ML Classification of Appendiceal Cancer

Team Manager — Senior Capstone

Wake Forest University · Aug 2020 — May 2021

Researched and developed ML models to process and classify histopathological images of appendiceal cancer — a rare, frequently misdiagnosed disease.
Built an object detection model using Faster-RCNN with Inception-ResNet-v2 to detect cancerous signet ring cells within histopathological images.
Developed a deep CNN using TensorFlow/Keras to classify appendiceal cancer subtypes with 85% accuracy vs. avg. pathologist accuracy of 52%.
Delivered a Flask/Python web app to Wake Forest Baptist Hospital doctors and pathologists for real-time image upload and model prediction.

0%Model accuracy

0%Avg. pathologist

#0Best capstone overall

View slides ↗Capstone deck

05 / Live Demo

See It
Think

This is the same idea behind the sentence-embedding work I do at Stanson Health — turning language into vectors a machine can compare. Except here, it's running entirely in your browser. No server, no API call — just a small model downloaded once and cached.

Runs fully client-side via transformers.js — nothing you type ever leaves your browser.

06 / Beyond Work

Outside
the Office

Ultramarathon Running 48th place — Colorado Run Rabbit Run 50 mile ultra

Triathlon Racing Triathlete — road & time trial

Backcountry Skiing AIARE Level 1 Certified

Mountain Biking Trail & enduro riding

SoundCloud · @cooleyethan

Open full profile on SoundCloud →

07 / Recent Activity

Recent
Miles

Pulled live from Strava — training totals for the year, my latest route, and my most recent workouts.

Last 5 Activities

Loading recent activity…

Open full profile on Strava →

08 / Contact

Get in
Touch

Interested in working together or just want to say hello? Send a message and I'll get back to you.

↗ linkedin.com/in/ethancooley

ThePerson

ProfessionalHistory

Stanson Health

TechnicalStack

AcademicBackground

FeaturedWork