Member of Technical Staff, Systems Infrastructure (2026 PhD New Grad)

The role

You will design and build core infrastructure systems including schedulers, storage systems, and networks for large-scale training and inference. You will model system behavior and turn research ideas into production systems.

What you will do

• Design, build, and operate core infrastructure systems for large-scale training and inference
• Model and measure system behavior using benchmarks, traces, and simulators
• Identify bottlenecks across the stack from kernel and driver to scheduler policy
• Turn research ideas into production systems
• Work closely with research and inference teams
• Contribute to the team's technical direction by tracking emerging hardware, interconnects, and systems research

Required qualifications

• PhD completed within the last 6 months, or expected completion by December 2026, in Computer Science, Computer Engineering, Electrical Engineering, or a similar field
• Research background in one or more of: distributed systems, operating systems, scheduling and resource management, storage systems, computer networks, computer architecture, or high-performance comput
• Depth in at least one of the three focus areas: Scheduling & resource management, Distributed storage & caching, or Datacenter networking
• Strong systems programming skills in C/C++, Rust, Go, or a similar language
• Working proficiency in Python
• Experience building and evaluating real systems and reasoning about performance with measurements
• Ability to communicate systems design and results clearly to audiences with different backgrounds

Nice to have

• First-authored publications at top-tier systems or networking venues (OSDI, SOSP, NSDI, EuroSys, SIGCOMM, ATC, FAST, ASPLOS, MLSys, SC, or similar)
• Industry internships in infrastructure, cloud, or HPC, or meaningful open-source contributions to systems projects (Kubernetes, Ray, Slurm, vLLM, NCCL, DPDK/SPDK, Ceph, or similar)
• Hands-on experience with GPU clusters, accelerator programming (CUDA, ROCm, Triton), or heterogeneous compute environments
• Experience with Kubernetes and cloud infrastructure, or with operating multi-tenant clusters in production
• Experience with performance profiling and debugging in distributed environments
• Research or engineering experience demonstrated via grants, fellowships, patents, or systems competitions

Benefits

• Solve Hard Problems
• Build What’s Next
• Ownership & Impact
• Learn from the Best

About the company

Fireworks is a Series D company valued at $17.5 billion that provides a platform for specialized intelligence, enabling companies to build, train, and serve AI models. The team was founded by the creators of PyTorch.