George Mastrapas

Athens, Attica, Greece
email-iconphone-icongithub-logolinkedin-logotwitter-logostackoverflow-logofacebook-logo
Join Prog.AI to see contacts
email-iconphone-icongithub-logolinkedin-logotwitter-logostackoverflow-logofacebook-logo
Join Prog.AI to see contacts

Summary

🤩
Rockstar
🎓
Top School
George Mastrapas is an ML Research Engineer based in Athens with seven years of experience building and productionizing machine learning and NLP systems. He blends research-minded experimentation with software engineering and MLOps practice, contributing backend and evaluation improvements to notable open-source projects such as jina-ai/finetuner for task-oriented embedding tuning. Comfortable refactoring evaluation infrastructure and exposing evaluator arguments, he focuses on making model evaluation clearer and more reproducible in real-world pipelines. With strong academic roots from the National Technical University of Athens, he pairs rigorous technical foundations with hands-on delivery across model development, deployment, and documentation.
code7 years of coding experience
book8.80, 8.80 at National Technical University of Athens
languagesGreek, English, Italian
github-logo-circle

Github Skills (12)

machine-learning10
python10
evaluation10
fine-tuning10
back-end-development9
jina9
transfer-learning8
pre-trained-model8
few-shot-learning7
pytorch6
docker6
dockers6

Programming languages (3)

ShellJupyter NotebookPython

Github contributions (5)

github-logo-circle
jina-ai/finetuner

Dec 2021 - Jan 2023

:dart: Task-oriented embedding tuning for BERT, CLIP, etc.
Role in this project:
userBack-end Developer & MLOps Engineer
Contributions:4 releases, 407 reviews, 23 commits in 1 year 1 month
Contributions summary:George primarily contributed to the Finetuner project by addressing issues related to the evaluator integration, including refactoring code for clearer implementation and adding evaluation features. They updated and refined the code related to model evaluation. Additionally, the user made minor documentation fixes and exposed arguments in the evaluator.
nlptriplet-lossfinetuningsiamese-networkjina
jina-ai/openclip

Jan 2024 - Nov 2024

An open source implementation of CLIP
Contributions:27 reviews, 47 PRs, 133 pushes in 9 months
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.
Request Free Trial
George Mastrapas