We are seeking a talented Reinforcement Learning (RL) Specialist to join our team for a game-related project. In this role, you’ll design, develop, and train AI agents using reinforcement learning techniques in the context of video games. Your work will directly contribute to technologies for automated testing, game balancing, and in-game AI. This is a unique opportunity to blend your expertise in RL with a passion for gaming, delivering solutions that elevate gameplay and streamline development processes.

Requirements

• Expertise in reinforcement learning, including experience training AI agents in simulated or real-world environments.  

• Proficiency in Python and familiarity with popular RL frameworks (e.g., Stable-Baselines, Ray/RLlib or similar).  

• Ability to write clean, efficient, and maintainable code.  

• Experience with machine learning concepts and workflows (e.g., reward shaping, exploration strategies, policy optimization).  

• Practical knowledge of Git.

• Ability to design clean code architecture.

• Ability to read and write documentation in English.

• Experience or willingness to learn Unreal Engine.

Nice to have

• Knowledge of C++ and experience optimizing RL models for performance (a strong plus).  

• Familiarity with game engines (e.g., Unity, Unreal Engine) and game development workflows.  

• Interest in video games, with an understanding of game design principles or mechanics.  

• Familiarity with Unreal Engine’s Learning Agents framework.

Your duties

• Design and implement reinforcement learning pipelines to train AI agents for video game environments.  

• Collaborate with Grail’s team to integrate AI agents into testing pipelines, balance game mechanics, and enhance player experiences in a diverse range of titles.  

• Optimize machine learning models for performance, scalability, and real-time application in games.  

• Contribute to the development of tools and workflows.

We offer

• Market-based salary adjusted to your skills and experience.

• Flexible working hours.

• Hybrid or remote working system.

• 10% of work time for self-development.

Submission instructions

• CV is required.

• Code samples are optional.

• Must be willing to complete a code test.