Summary
Mustafa Isik is a research engineering leader with 11 years of experience specializing in diffusion models, neural rendering, digital humans and 3D graphics. Now managing research engineering at Synthesia, he has repeatedly shipped real-time neural renderers and ultra-fast facial trackers, including a 650 FPS 3DMM tracker and the HumanRF system showcased at SIGGRAPH. He combines deep research instincts with hands-on systems engineering—writing optimized CUDA/Triton kernels, integrating OptiX, and deploying models with TensorRT for state-of-the-art runtime performance. His background spans academia and industry with graduate work at TUM and ETH Zürich and impactful internships at Adobe (SIGGRAPH-published and patented denoising work). Known for modernizing interdependent codebases and building production data pipelines for large-scale video training, he bridges cutting-edge research and practical product delivery. Based in London, he favors projects that make digital humans more expressive and real-time capable, often squeezing extreme performance from constrained compute.
11 years of coding experience