Mikkel Garcia
Software Engineer at 255LABS
Denver, Colorado, United States
Summary
🤩
Rockstar🎓
Top SchoolMikkel Garcia is a Denver-based software engineer with 16 years of experience building scalable systems for startups, from Rails APIs powering kiosk networks to backend and ML work at 255LABS. He combines hands-on backend engineering and machine learning—authoring open-source ML tooling such as contributions to the HyperGAN project—with a pragmatic, technology-agnostic approach to solving real market needs. Mikkel has shipped cross-platform apps, introduced production-grade observability and testing practices, and helped multiple teams find product-market fit and scale. He’s an early adopter and presenter of emerging technologies (node, Docker, RNNs/GANs) and focuses on effectiveness and utility rather than trends. Notably, his work blends heavy technical contributions (custom GAN loss functions and gradient experimentations) with product-oriented thinking that drives startup growth.
16 years of coding experience
9 years of employment as a software developer
Bachelor of Science (B.S.), Computer Science, Bachelor of Science (B.S.), Computer Science at Texas Tech UniversityUniversity of Colorado Colorado Springs
Stackoverflow
61reputation
2kreached
3answers
0questions
Github Skills (15)
machine-learning10
loss-functions10
gradient-descent10
tensorflow10
cgan10
cyclegan10
python10
dcgan10
deep-learning9
pytorch9
facebook-graph-api6
facebook-open-graph6
regex6
instagram6
facebook6
Programming languages (4)
Github contributions (5)
HyperGAN/HyperGAN
Jun 2016 - Jan 2023
Composable GAN framework with api and user interface
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:1 release, 1981 commits, 113 PRs in 6 years 7 months
Contributions summary:Mikkel implemented gradient-related experimentations, likely related to GAN training, including the development of gradient-based optimizations and analysis. Their work involved adding functionality for a gradient-based loss function (gradient penalty) and creating components for a modified GAN architecture. They also introduced a new, custom loss function and related metrics, which required integration with existing components and consideration of the original loss function.
pytorchapiuser-interfacedeep-learninggenerative-adversarial-network
HyperGAN/HyperGAN-tutorials
Jul 2019 - Aug 2019
Learn how to use hypergan in your games, apps and websites.
Contributions:24 commits, 2 PRs, 22 pushes in 8 days
gamegames
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.