What you will do:

• Build and train machine learning models that replicate football players’ behavior in a simulated or game-like environment;

• Apply Imitation Learning, Reinforcement Learning, Decision Transformers, and their hybrids to model realistic and dynamic agent behavior;

• Design and maintain data pipelines for training, validation, testing, and inference of ML models;

• Contribute to the architecture of an AI system that will evolve from MVP stage to full-scale deployment;

• Run behavior simulations to test models, evaluate realism and tactical consistency, and optimize learning approaches;

• Collaborate with game designers and engineers to bring AI-driven behavior into real-time environments (e.g., via API or engine integration).

Skills and requirements: 

• Solid experience with PyTorch or TensorFlow, especially in sequence modeling or reinforcement/imitation learning use cases;

• Hands-on expertise in one or more RL algorithms (e.g., PPO, SAC, DQN) and at least one IL method (e.g., Behavioral Cloning, Inverse Reinforcement Learning);

• Good grasp of sequence modeling techniques, including RNN/LSTM and Transformers, particularly Decision Transformers;

• Experience in training from offline datasets - such as time-series or trajectory logs (e.g., player position tracking);

• Familiarity with modern ML toolchains: config management (e.g., Hydra), experiment tracking (Weights & Biases), and data versioning (DVC);

• Ability to build inference-ready pipelines (e.g., using ONNX, TorchScript, or REST API) and interface them with external systems.

Nice-to-Have:

• Prior experience integrating ML models into game engines (Unreal / Unity etc);

• General knowledge of football (ideally - tactical roles, and/or player behavior modeling);

• Exposure to offline RL, GFlowNets, or World Models;

• Contributions to open-source projects or participation in AI-related competitions (e.g., NeurIPS, Kaggle, AIcrowd);

• Strong debugging, visualization, or model interpretability skills.

Why choose us?

• The opportunity to shape AI architecture in a project that blends gaming, simulation, and machine learning;

• Full ownership of core technical direction in AI behavior modeling and training;

• Collaboration with a small, focused team working on a next-gen product from early-stage design.
  
  We offer:

• Competitive salary;

• Medical insurance, sick leaves and social benefits;

• Relocation package;

• Sports compensation;

• Online English classes;

• Referral bonuses;

• Co-working compensation;

• Corporate events.

  
  Please send your CV using the form below to apply for this opportunity.