Senior ML Systems Engineer at well-funded media AI startup

You will build the foundational infrastructure that powers next-generation filmmaking technology. This role focuses on designing high-performance data platforms and training systems for TB-scale multimodal datasets. You will bridge the gap between research and production, ensuring distributed training pipelines and inference services are efficient, reliable, and scalable for Hollywood-grade content.

Well-funded media AI startup

• Build and optimize a multimodal data platform using tools like Arrow, Parquet, and vector search to curate massive video datasets.
• Scale distributed training pipelines using PyTorch and Ray to improve performance across multi-node GPU setups.
• Own the production inference stack, optimizing model ensembles and request protocols using Triton for high-throughput delivery.

• Proven track record building ML infrastructure, training platforms, or data systems rather than just training models.
• Deep expertise in Python systems engineering and hands-on experience with PyTorch internals and distributed training (DDP/FSDP).
• Strong intuition for data performance trade-offs and experience handling TB-scale datasets or high-throughput media pipelines.

• Work at the intersection of generative AI and entertainment, building tools that give filmmakers creative superpowers.
• Join a well-funded team backed by top-tier VCs where ML infrastructure is the core product, not a cost center.
• Solve complex technical challenges involving large-scale video data, high-performance compute clusters, and real-time inference.