Applied ML Engineer

About the Role

We’re looking for an Applied ML Engineer to turn cutting-edge video generation research into production-ready systems. You’ll take the latest ideas from papers, open-source models, and APIs and make them real — implementing, optimizing, and scaling them into fast, reliable user-facing workflows. Unlike a research-only role, this position is focused on engineering execution: building, testing, and refining generative video models so they run efficiently and deliver great user experiences. You’ll work closely with researchers to translate prototypes into robust ML pipelines, and with product engineers to integrate them into interactive features at scale.

What You’ll Do

• Implement and optimize state-of-the-art models (diffusion, transformers, autoregressive, hybrid) for training and inference.
• Productionize prototypes: adapt academic/OSS code into reliable, well-structured, and maintainable systems.
• Optimize performance: tune models for latency, throughput, memory efficiency, and cost across GPU/accelerator environments.
• Build applied workflows on top of foundation models — controllable video editing, personalization, real-time interaction.
• Benchmark rigorously: measure tradeoffs in fidelity, controllability, latency, and scalability, and feed learnings back into the product.
• Collaborate cross-functionally with researchers (for new methods) and product engineers (for integration and UX).
• Maintain high engineering standards: clean code, reproducible experiments, robust pipelines, and deployment practices.

What We’re Looking For

• Strong background in applied machine learning or ML engineering , ideally in vision, video, or multimodal ML.
• Experience implementing and optimizing diffusion, transformer-based, or autoregressive models.
• Proficiency with PyTorch and modern ML tooling (training, inference, profiling, distributed training/inference).
• Proven ability to translate research code into stable, production-grade systems.
• Focus on applied outcomes: making ML systems fast, interactive, and scalable for real users.
• Bonus: experience with video editing, inpainting, or interactive media systems.
• Bonus: experience with deployment at scale (Kubernetes, Triton, ONNX, quantization, distillation).

Why Join Us

We’re building a new kind of media company — one that combines creative freedom with cutting-edge ML. You’ll join a team of ex-founders, ICML-published researchers, and builders from unicorn startups and top UGC platforms, backed by a16z Speedrun and top angels in the video AI space. We’re still early, and this is your chance to shape both the technology and the culture from day one.