**Summary:** The MRS ML Infra team is focusing on ML Infra performance and efficiency for both large scale AI training and inference workflows in the recommendation domain.In this role, you will work on optimizing the e2e stack for model training and inference for large scale recommendation models, with opportunities coming from the domains of distributed systems, model/system co-design, GPU optimizations, and more. While the core of day-to-day work and key responsibility will be to identify and lead the execution for short/mid term opportunities for efficiency optimization, you will also drive long term strategies and shape team direction on things like model/system co-design, performance automation, regression detection and mitigation, etc. **Required Skills:** Software Engineer, Infrastructure Responsibilities: 1. Identify performance opportunities and bottlenecks across a wide range of MRS models, infrastructure and systems. 2. Implement changes to capture efficiency improvements. 3. Guide other engineers both inside and outside the team to execute on efficiency and performance opportunities, issues and bottlenecks. 4. Drive cross-functional collaborations and alignments with multiple partner or product ML teams. 5. Define technical direction(s), strategy and roadmap for the team. 6. Provide mentorship and guidance to grow other teammates. **Minimum Qualifications:** Minimum Qualifications: 7. BS/MS in Electrical Engineering, Computer Science or a related field or equivalent experience. 8. 5+ years of experience in AI Infra or System performance. 9. Hands-on experience in optimizing complex software solutions, such as distributed systems, large scale CPU/GPU clusters, or similar. 10. Demonstrated experience in driving team execution and reaching alignment with cross-functional partners 11. Previous experience in mentoring and growing software and/or machine-learning engineers as either a tech lead or a manager. 12. Capacity to investigate and debug issues in complex systems, including ones spanning multiple components or sub-systems **Preferred Qualifications:** Preferred Qualifications: 13. Hands-on experience with large-scale AI infra systems (for example, GPU training clusters) 14. Experience in training and/or inference solutions for large models (e.g. recommendation models or LLMs). 15. Experience in high performance computing including communication optimization, CUDA kernel optimization, distributed training and inference, etc. **Industry:** Internet