We’re looking for a senior Unreal Engine 5 developer to build the technical integration layer connecting AI models to MetaHuman characters. The focus is on the systems and architecture that pipe AI model outputs into UE5 — not on creating MetaHumans or voice content itself.
Responsibilities
∙ Architect and build the integration layer between AI model APIs and UE5 runtime systems
∙ Build async C++ subsystems to query AI models and route outputs to the appropriate UE5 systems
∙ Develop the data pipeline that translates AI model responses into UE5-readable inputs (animation params, behavior triggers, state changes)
∙ Implement plugin architecture so AI models can be swapped or updated without engine changes
∙ Build robust error handling, fallback logic, and latency management for live model inference
∙ Create Blueprint-exposed interfaces so the team can wire AI outputs to MetaHuman systems without C++
∙ Profile and optimize the integration layer to keep inference calls from impacting frame rate
Requirements
∙ 5+ years of professional Unreal Engine experience
∙ Deep C++ skills — custom subsystems, async task systems, UE plugin development
∙ Proven experience integrating external APIs or AI models into a real-time engine
∙ Strong understanding of UE5 game loop, threading model, and performance constraints
∙ Experience designing clean interfaces between engine systems and external services
∙ Familiarity with model inference concepts: latency, async calls, context management
Nice to Have
∙ Experience with model serving frameworks (Ollama, TensorRT, OpenAI API, Hugging Face)
∙ Knowledge of UE5 animation systems sufficient to understand downstream consumers of the data
∙ Local model inference experience (llama.cpp, ONNX Runtime)
∙ Prior work building engine-level tooling or editor plugins
Also, must show specific examples related to this. Only senior people should be applying.
... Show more