Principal AI Engineer

Principal AI Engineer - Reinforcement Learning & Game AI

Dubai - Onsite

Are you a thinker and a builder?
We're looking for an exceptional Principal AI Engineer to design and build the next generation of poker and game AI from the ground up - redefining how strategy-based games are learned and played.

This isn't about maintaining existing systems - it's about creating something entirely new. You'll architect, experiment, and deliver pioneering work in reinforcement learning, game theory, and AI performance optimization.

What You'll Do

• Design and build new AI systems for imperfect information games like poker (NLH, PLO, and variants).
• Lead the C++ architecture and development of a high-performance AI engine.
• Research and implement advanced reinforcement learning and game theory techniques.
• Collaborate closely with senior leadership to turn bold ideas into practical, high-impact features.
• Publish, present, and gain visibility as a domain expert in poker AI and imperfect information games.

What You'll Bring

• Poker expertise (non-negotiable): strong understanding of GTO concepts, solvers, and poker theory.
• C++ mastery: proven track record building complex, high-performance systems.
• Reinforcement Learning experience (3+ years).
• Strong knowledge of game theory, neural networks, and RL algorithms (CFR, PPO, A3C, MCTS, etc.).
• Builder mindset - thrive on turning ideas into working systems from a blank slate.
• Bonus: experience with multi-agent RL, published research, or open-source AI projects.

What's in it for you?

• 60-80k Dirham per month
• Real impact - shape the future of AI in strategic gaming
• Freedom to innovate - explore, experiment, and publish
• Cutting-edge work - push the limits of reinforcement learning and game AI
• Recognition - opportunities to present and publish your work globally.
• Collaborative culture - work alongside passionate engineers and researchers
• Relocation package

If you're passionate about AI, poker, and solving the toughest problems in imperfect information games, we'd love to hear from you.