Irismart Recruiting

The AI systems you build are the game's core differentiator - not a feature, not a tech demo. 
If you've ever played an MMO and thought "this world feels scripted and dead" and had opinions about how to fix it, this is the job.

You will own the design, development, and gameplay integration of all three AI systems, described above. 

Your deliverables are systems that players feel, not documents or dashboards.

What you'll be doing:

● Design and train models for world scenario simulation and player impact analysis.

● Build the translation layer that converts AI decisions (text / parametric) into clean, engine-ready data for UE5.

● Integrate AI outputs into gameplay loops so events and world reactions feel like organic game mechanics - not bolted-on AI demos.

● Work closely with Game Designers and the CTO to ensure AI pacing respects game feel: tension, rhythm, player agency.

Requirements:

● Strong Python (LLM orchestration, agent frameworks - LangChain/LangGraph or equivalent).

● Experience building multi-agent systems with defined roles, memory, and inter-agent communication.

● Solid understanding of LLM fine-tuning, prompt engineering, and RAG pipelines.

● First-hand gaming - either professional gamedev background or engagement with complex games (MMO, survival, strategy - you understand what "game feel" means from the inside).

● Ability to define and maintain a world state model that multiple agents read from and write to.