AI/ML Engineer

Remote | up to 1300 RUB/hour

Hello everyone!

For the project, we are looking for a strong AI/ML engineer for the FaceFusion GPU Pipeline. This is a key role in the project: you will need to launch an AI video face replacement pipeline from scratch and achieve a speed of 5-15 seconds per clip (on T4/A10G GPUs).

Hard requirements (Must have):

• Code: Python from 2.5 years - strictly production code (not jupyter notebooks).
• ML/Inference: ONNX Runtime (loading, CUDA EP, FP16/INT8 quantization).
• Computer Vision: practical experience with face detection, face alignment, landmarks.
• Hardware: NVIDIA GPU - experience working with the CUDA stack in production.
• Infrastructure: writing REST API microservices on FastAPI, building and optimizing GPU containers in Docker.

Will be a huge plus (will set you apart from everyone):

• Direct experience working with the FaceFusion framework.
• ArcFace embeddings.
• TensorRT: engine compilation and optimization.
• NVENC (FFmpeg hardware encoding).
• Kubernetes + GPU node pool.

Conditions:

• Complex engineering challenges, modern stack without legacy.
• Project-based employment.
• Payment within the hourly rate of up to 1300 RUB/hour (determined based on interview results).
• Mandatory possession of an Individual Entrepreneur (IP) status or self-employed status in the Russian Federation.
• Important: A small test task is provided for candidates.

Outstaff studios and agencies are asked not to disturb!

Contact Glebque