Open to opportunities

Justin Barry

@justinbarry

Message

ML research engineer building production-ready deep learning systems across vision, language, and agents.

United States

Message

What I'm looking for

I seek roles where I can design end-to-end ML systems, tackle ambiguous problems, lead research-to-production work, and mentor teams in an experimentation-driven culture.

I am an ML research engineer and applied scientist who designs and implements production-ready deep learning systems, owning both the math and PyTorch code. I work across vision, language, and structured data, adapting architectures from first principles to solve messy, ambiguous problems.

I have industry experience shipping models and experiments at scale, including work at Amazon Prime Video optimizing artwork selection with contextual bandits and two-tower recommendation models, and prototyping LLM-based ideation and multi-judge evaluation systems at Spotter. I build agentic AI pipelines and publish thorough, educational implementations and explainers of models like GPT, Stable Diffusion XL, Rectified Flow, CLIP, ViT, ResNet, UNet, VAEs, and AutoEncoderKL.

I advise teams end-to-end on problem framing, data strategy, modeling, experimentation, and deployment, and I contribute open-source projects and long-form technical content to share reproducible, production-ready solutions and teaching material.

Experience

Work history, roles, and key accomplishments

Current

ML Research Engineer

Current

Independent

Jan 2024 - Present (2 years)

Advise teams on end-to-end ML system design and build production-ready PyTorch models; publish first-principles implementations and educational content on generative and discriminative architectures.

PyTorch Modelling Generative Models Continuous Normalizing Flows Stable Diffusion Experimentation Deployment

Senior Machine Learning Scientist

Spotter

Jan 2023 - Jan 2024 (1 year)

Prototyped an LLM-based ideation and ranking system for YouTube creators and developed silver-label dataset pipelines and GPT-4–based judges to evaluate ideas and thumbnails; fine-tuned SDXL Turbo for thumbnail generation and benchmarked variants with human rubrics.

LLM Prompt Engineering Fine Tuning Évaluation A B Testing Model Evaluation

Senior Machine Learning Scientist

Independent

Jan 2022 - Jan 2023 (1 year)

Built predictive ML systems for UFC fights using scraped historical data and custom feature engineering to model match outcomes and performance.

Scraping Model Evaluation Python Experimentation Deployment

Machine Learning Scientist

Amazon

Jan 2019 - Jan 2022 (3 years)

Improved global streaming by 2.9% by optimizing cover art selection with contextual multi-armed bandits and deep models; built two-tower image–customer recommendation models and large-scale PySpark Bayesian logistic regression for inference.

Deep Learning Recommender Systems Pyspark Representation Learning LDA Wasserstein Autoencoders Experiment Design

Senior Software Engineer

CSRA Inc

Jan 2013 - Jan 2017 (4 years)

Led enterprise Java application development and Docker-based testing for federal clients, focusing on reliability, security, and prototyping.

Java Docker enterprise applications Testing Prototyping Reliability & Security CI CD

Software Engineer

Booz Allen Hamilton

Jan 2011 - Jan 2013 (2 years)

Built secure enterprise systems, REST services, and internal tools for government customers with emphasis on secure coding and service design.

REST APIs Enterprise Systems Tooling Java Systems Design Integration Testing

Research Intern

Thomas Jefferson National Accelerator Facility

Jan 2010 - Jan 2011 (1 year)

Developed a multithreaded histogram data aggregation system for particle physics experiments to support high-throughput data processing.

Multithreading Data Aggregation High Throughput Systems C C++Scientific Computing Performance Optimization

Education

Degrees, certifications, and relevant coursework

National GEM Consortium

GEM Fellow, Fellowship

2018 - 2019

GEM Fellowship recipient supporting graduate study and professional development in engineering and science.

University of Central Florida

Master of Science, Computer Science

2017 - 2019

Activities and societies: GEM Fellowship (National GEM Consortium) 2018–2019

Completed a Master of Science in Computer Science while on a former PhD track, focusing on machine learning and related research topics.

Christopher Newport University

Bachelor of Science, Mathematics and Computer Science

2006 - 2011

Earned a dual Bachelor of Science degree in Mathematics and Computer Science with coursework supporting algorithms, statistics, and software development.