Summary
Prashanth Thinakaran is an AI infrastructure engineer based in Palo Alto with 11 years of experience building and scaling GPU-first clusters for large models and latency-sensitive workloads. He brings deep HPC and cloud systems expertise from a PhD in computer science and hands-on roles at Cerebras, KLA, and GPU-focused startups, where he designed automation, networking (RoCEv2), and scheduler integrations for SLURM and Kubernetes. At Cerebras he helped scale wafer-scale systems from petaflops to exaflops for 70B–175B parameter models, and at a stealth GPU cloud he led customer-facing solutions engineering for HGX-H100 fleets. Prashanth is adept at triaging production reliability across hardware, networking, and storage boundaries, and he translates customer pain points into product roadmap decisions. He combines research-grade thinking about heterogeneity-aware scheduling (from his dissertation) with pragmatic tooling—Ansible, Python, and internal ops automation—to deliver repeatable deployments. Colleagues rely on him as cross-functional glue who turns academic insights into production-grade AI infrastructure.
11 years of coding experience
9 years of employment as a software developer
Doctor of Philosophy (Ph.D.) Computer science and engineering, Doctor of Philosophy (Ph.D.) Computer science and engineering at Penn State University
Bachelor of Technology Information Technology, Bachelor of Technology Information Technology at Anna University Chennai
English, Tamil