AI Performance Optimization Engineer

<div class="content-intro"><h1><strong>Who We Are</strong></h1> <p>Lightning AI is the company reimagining the way AI is built. After creating and releasing PyTorch Lightning in 2019, Lightning AI was launched to reshape the development of artificial intelligence products for commercial and academic use.</p> <p>We are on a mission to simplify AI development, making it accessible to everyone—from solo researchers to large enterprises. By removing the complexity of building and deploying AI tools, we empower innovators to focus on solving real-world problems. Our platform is built to scale with the latest AI advancements while staying intuitive and adaptable, so you can bring your ideas to life.</p> <p>We have offices in New York City, San Francisco, and London and are backed by investors such as Coatue, Index Ventures, Bain Capital Ventures, and Firstminute.</p> <div class="c-message_actions__container c-message__actions"> <h1 class="c-message_actions__group" data-qa="message-actions"><strong>Our Values</strong></h1> </div> <ul> <li> <p><strong>Move Fast</strong>: We act with speed and precision, breaking down big challenges into achievable steps.</p> </li> <li> <p><strong>Focus</strong>: We complete one goal at a time with care, collaborating as a team to deliver features with precision.</p> </li> <li> <p><strong>Balance</strong>: Sustained performance comes from rest and recovery. We ensure a healthy work-life balance to keep you at your best.</p> </li> <li> <p><strong>Craftsmanship</strong>: Innovation through excellence. Every detail matters, and we take pride in mastering our craft.</p> </li> <li> <p><strong>Minimal</strong>: Simplicity drives our innovation. We eliminate complexity through discipline and focus on what truly matters.</p> </li> </ul></div><h2>What We're Looking For</h2> <p>We are seeking a highly skilled AI Optimization Engineer to work on optimizing training and inference workloads on compute accelerators and clusters, through  the <strong>Lightning Thunder compiler</strong> and the broader <strong>PyTorch Lightning ecosystem</strong>. This role sits at the intersection of <strong>deep learning research, compiler development, and large-scale system optimization</strong>. You’ll be shaping technology that pushes the boundaries of model performance and efficiency, creating foundational software that will impact the entire machine learning ecosystem.<br><br>You will be joining the Engineering Team and report to our Tech Lead. This is a hybrid role based in either our New York City or San Francisco office with in-office requirements of 2 days per week. The salary range for this role is $120,000-$250,000. </p> <h2><strong>What you’ll do</strong></h2> <ul> <li><strong>Develop performance-oriented model optimizations</strong> at multiple levels:</li> <ul> <li>Graph-level (e.g., operator fusion, kernel scheduling, memory planning)</li> <li>Kernel-level (CUDA, Triton, custom operators for specialized hardware)</li> <li>System-level (distributed training across GPUs/TPUs, inference serving at scale)</li> </ul> <li><strong>Advance the Thunder compiler</strong> by building optimization passes, graph transformations, and integration hooks to accelerate training and inference workloads.</li> <li><strong>Work across the software stack</strong> to ensure optimizations are accessible to end users through clean APIs, automated tooling, and seamless integrat ... (truncated, view full listing at source)