Senior Generative Video ML Engineer, Post-Training & Fine-Tuning
Core
Senior ML engineer responsible for post-training, fine-tuning, and reproducible inference of open-weight generative video models to ensure character, scene, and temporal consistency.
Role type
Senior hands-on IC machine-learning engineer (generative video)
Builds
Reproducible inference environments, post-training adaptations (LoRA, adapters, supervised tuning), and evaluation pipelines for video generation models.
Domain
Generative AI, Video Production, Media Rights & Compliance
Deliverable
production ML models
Required skills
Diffusion/flow-based video generation, Transformer-based video models, PyTorch, GPU systems (distributed training, memory optimization), Experiment tracking, Evaluation pipelines (identity, temporal quality)
Preferred skills
Wan, MiniMax, HunyuanVideo, CogVideoX, LTX Video, Mochi, Character-consistency techniques, Reference conditioning, Identity embeddings, H100/H200 optimization, Blinded human review design
Technologies
PyTorch, LoRA, Adapters, H100, H200
Responsibilities
Stand up reproducible inference environments, Design and run license-permitted post-training experiments, Measure identity consistency and temporal stability, Diagnose failures and compare remediation strategies, Package code and technical handover materials
Seniority
Senior, hands-on IC
