Tech Lead, AI Compute Infrastructure

<div class="content-intro"><p><strong>About HeyGen</strong></p> <p>At HeyGen, our mission is to make visual storytelling accessible to all. Over the last decade, visual content has become the preferred method of information creation, consumption, and retention. But the ability to create such content, in particular videos, continues to be costly and challenging to scale. Our ambition is to build technology that equips more people with the power to reach, captivate, and inspire audiences.<br>Learn more at <a href="https://www.heygen.com">www.heygen.com</a>. Visit our Mission and Culture doc <a href="https://docs.google.com/document/d/1JQ6k6Ce8VRE60zO09Uc94JVaxRWA2xC1-IbIFAAHPvs/edit?usp=sharing">here</a>. </p></div><p>We are seeking a seasoned<strong> Technical Leader</strong> to build and scale the foundational compute infrastructure that powers our state-of-the-art AI models—from multimodal training data pipelines to high-throughput, low-latency video generation.</p> <h3>Responsibilities</h3> <p>You will be the core engineer responsible for building the robust, efficient, and scalable platform that enables our research and production teams to rapidly iterate on HeyGen's generative video models. Your contributions will directly impact model performance, developer productivity, and the final quality of every AI-generated video.</p> <ul> <li> <p><strong>Optimize GPU Utilization:</strong> Design and implement mechanisms to aggressively <strong>optimize GPU and cluster utilization</strong> across thousands of devices for inference, training, data processing and large-scale deployment of our <strong>state-of-art video generation models</strong>.</p> </li> <li> <p><strong>Develop Large-Scale AI Job Framework:</strong> Build highly scalable, reliable frameworks for launching and managing massive, heterogeneous compute jobs, including multi-modal high-volume data ingestion/processing, distributed model training, and continuous evaluation/benchmarking.</p> </li> <li> <p><strong>Enhance Observability:</strong> Develop world-class <strong>observability, tracing, and visualization tools</strong> for our compute cluster to ensure reliability, diagnose performance bottlenecks (e.g., memory, bandwidth, communication).</p> </li> <li> <p><strong>Accelerate Pipelines:</strong> Collaborate closely with AI researchers and AI engineers to integrate innovative acceleration techniques (e.g., custom CUDA kernels, distributed training libraries) into production-ready, scalable training and inference pipelines.</p> </li> <li> <p><strong>Infrastructure Management:</strong> Champion the adoption and optimization of modern cloud and container technologies (<strong>Kubernetes, Ray</strong>) for elastic, cost-efficient scaling of our distributed systems.</p> </li> </ul> <h3>Minimum Requirements</h3> <p>We are looking for a highly motivated engineer with deep experience operating and optimizing AI infrastructure at scale.</p> <ul> <li> <p>Bachelor's degree in Computer Science, Engineering, or a related field, or equivalent practical experience.</p> </li> <li> <p>5+ years of full-time industry experience in <strong>large-scale MLOps, AI infrastructure, or HPC systems</strong>.</p> </li> <li> <p>Experience with data frameworks and standards like <strong>Ray, Apache Spark, LanceDB</strong></p> </li> <li> <p>Strong proficiency in <strong>Python</strong> and a high-performance language such as <strong>C++ </strong> for developing core infrastructure components.</p> </li> <li> <p>Deep understanding and hands-on experience with modern orchestration and distributed computing frameworks such as <strong>Kubernetes and Ray</strong>.</p> </li> <li> <p>Experience with core ML frameworks such as <strong>PyTorch, TensorFlow, or JAX</strong>.</p> </li> </ul> <h3>Preferred Qualifications</h3> <ul> <li> <p>Master's or PhD in Computer Science or a related technical field.</p> </li> <li> <p>Demonstrated Tech Lead experience, driving projects from conceptual design through to pro ... (truncated, view full listing at source)