Milan Gritta
AI Research Engineer at Tether.io
London, England, United Kingdom
Summary
👤
Senior🎓
Top SchoolMilan Gritta is an AI research engineer with a decade of experience bridging computational linguistics, applied NLP research, and product-focused engineering. He holds a PhD from Cambridge and has driven R&D in conversational AI, multilingual NLP and code-generation LLMs at Huawei before moving to a startup-style role at Tether.io. Milan combines meticulous scientific methodology with practical engineering—contributing training pipelines and alignment work to notable open projects like Huawei Noah Research—so his solutions are both reproducible and production-ready. He is comfortable across the research-to-product lifecycle, from data generation and augmentation to model prototyping and deployment. Early experience in customer-facing roles gives him a pragmatic user-centered perspective that informs his approach to building useful language systems.
10 years of coding experience
7 years of employment as a software developer
Doctor of Philosophy (Ph.D.) Computational Linguistics, Doctor of Philosophy (Ph.D.) Computational Linguistics at University of CambridgeBachelor of Science (BSc) Computer Science, Bachelor of Science (BSc) Computer Science at University of Sussex
German, Slovak, English
Github Skills (6)
transformers10
pytorch10
machine-learning10
nlp10
python10
cross-entropy8
Programming languages (3)
Github contributions (5)
huawei-noah/noah-research
Aug 2021 - May 2022
Noah Research
Role in this project:
ML Engineer
ML EngineerContributions:21 commits, 3 PRs, 20 pushes in 8 months
Contributions summary:Milan's commits primarily involve modifications to the `train.py` file within the `xero_align` directory, suggesting a focus on model training and related functionalities. Code additions include a `Trainer` class and related training loops, indicating the development of machine learning training pipelines. The inclusion of various losses such as MSELoss, and CrossEntropyLoss along with the use of a model’s roberta implementation points toward the user's involvement in building and training models with a focus on language alignment.
noah
milangritta/WhatsMissingInGeoparsing
Mar 2016 - Oct 2019
Contributions:35 commits, 34 pushes, 1 branch in 3 years 7 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.