Miljan Martic
London, England, United Kingdom
Summary
🤩
Rockstar🎓
Top SchoolMiljan Martic is an ML-focused software engineer and founder with 11 years of experience applying advanced AI to web3 and real-world systems. He spent several years at DeepMind as a Research Engineer and Senior Research Engineering Team Lead, contributing to widely used reinforcement learning projects like TRFL and ai-safety-gridworlds where he improved core RL tests and implemented new environments. Miljan combines hands-on research engineering with product-facing experience as Co-Founder & CEO of Kosen Labs, and earlier roles across adtech and finance where he built RL-driven optimization and credit-risk models. He holds top-ranked degrees in computer science and machine learning (UCL MSc, University of Manchester, University of Belgrade) and is comfortable shipping robust tests, infrastructure, and novel environment logic—work that often sits between research prototypes and production-ready systems.
11 years of coding experience
8 years of employment as a software developer
BSc (Hons), Computer Science, First class (82%), BSc (Hons), Computer Science, First class (82%) at The University of ManchesterBSc (Hons), Information Systems and Technologies, 9.0/10.0, BSc (Hons), Information Systems and Technologies, 9.0/10.0 at University of BelgradeUniversity College London
English, Serbian, Croatian, Russian, Spanish
Github Skills (11)
environmental10
dev-environment10
game-development10
tensorflow10
environ10
python10
reinforcement-learning10
enviroment10
testing10
pytest9
machine-learning9
Programming languages (2)
Github contributions (5)
google-deepmind/ai-safety-gridworlds
Nov 2017 - Oct 2020
This is a suite of reinforcement learning environments illustrating various safety properties of intelligent agents.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:10 commits, 9 PRs, 13 pushes in 2 years 11 months
Contributions summary:Miljan contributed significantly to the `ai-safety-gridworlds` repository by introducing and modifying game environments. Their primary contribution involved implementing the "conveyor_belt" environment, including its variants, demonstrating their ability to create and modify game logic. Further work included creating a test file for the conveyor belt environment, showcasing the ability to write and integrate unit tests. These contributions suggest a focus on developing and extending the core functionality of the gridworlds project.
intelligent-agentsreinforcement-learningpropertiesagentsdeep-reinforcement-learning
google-deepmind/trfl
Jun 2018 - Dec 2019
TensorFlow Reinforcement Learning
Role in this project:ML Engineer
ML EngineerContributions:8 commits, 2 PRs, 8 pushes in 1 year 5 months
Contributions summary:Miljan contributed internal changes and updates to the TensorFlow Reinforcement Learning (TRFL) library. They modified test files, including those for indexing and distribution value operations, to check for shape errors and runtime behavior. Further contributions involved updates to the setup configuration file and the creation of new test files for value and retrace operations. These changes suggest a focus on improving the functionality and testing of core TRFL components.
reinforcement-learningreinforcementdeep-reinforcement-learningtensorflow
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.