Overview
The Foundry and Inference Training FIT team within Microsoft AI Infrastructure builds the platforms, systems, and operating mechanisms that enable teams to develop, train, deploy, and run AI services at scale. The FIT Infrastructure team builds storage and accelerated-compute platforms for large-scale AI training and inference on Azure. You will work across distributed storage, Kubernetes, GPU clusters, networking, and cloud-scale systems to deliver secure, reliable, high-performance, and cost-efficient infrastructure. As a Principal Software Engineer, Storage Infrastructure, you will design, build, and operate storage platforms for foundation-model pre-training, post-training, fine-tuning, and inference. You will partner with Azure infrastructure, networking, operating systems, and AI platform teams to solve challenging problems in storage performance, reliability, scalability, and developer productivity. This role offers opportunities to shape platform architecture, deepen expertise in distributed systems, and grow your technical leadership through broad cross-team impact. Microsoft’s mission is to empower every person and every organization on the planet to achieve more, and we’re dedicated to this mission across every aspect of our company. Our culture is centered on embracing a growth mindset and encouraging teams and leaders to bring their best each day. Join us and help shape the future of the world.
Responsibilities
Responsibilities:
- Design, build, and evolve distributed storage systems and file systems for cloud-scale AI training and inference workloads.
- Architect and operationalize storage stacks for large Kubernetes clusters, optimizing data access, throughput, latency, scalability, and reliability for GPU-accelerated workloads.
- Design storage and KVkey-value caching architectures that improve LLM inference latency, throughput, resource utilization, and cost efficiency.
- Improve durability, availability, performance, and operational maturity through automation, observability, diagnostics, incident response, and resiliency engineering.
- Analyze complex cross-layer production issues, identify root causes, and implement durable improvements across storage, compute, networking, and runtime components.
- Partner with infrastructure and machine learningLLM platform teams to deliver end-to-end capabilities for foundation-model pre-training, post-training, fine-tuning, and inference.
- Lead architecture reviews, influence long-term storage strategy, mentor engineers, and establish engineering practices for secure and reliable cloud services.
- Embody our Culture and Values
Qualifications
Required/minimum qualifications
- Bachelor's Degree in Computer Science or related technical field AND 6+ years technical engineering experience with coding in languages including, but not limited to, C, C++, C#, Java, JavaScript, or Python- OR equivalent experience.
