Aspi Siganporia ^{Ultimate Decision-Maker}

Senior Director of Engineering at Google

Current Job Duration: 4 mos (until March 2026)

HEADLINE:
Engineering Director

PERSONAL SUMMARY:
As the Engineering Director of ML Performance and Developer Tools, Aspi oversees the entire ML computing systems development lifecycle. His responsibilities include co-designing ML accelerators with hardware teams and optimizing ML workloads after systems are deployed in production.

Aspi's team is dedicated to building ML computing platforms that empower ML engineers to develop industry-leading AI/ML applications with high model quality, while simultaneously enhancing performance and reducing the total cost of ownership (TCO) with each new chip generation.

Performance Optimization: This involves increasing throughput, reducing latency in training step time, and decreasing serving latency, all while maintaining excellent model quality.

Developer Tools: These tools are designed to boost engineering productivity by helping ML developers understand the execution of their ML model training workloads. This understanding enables them to optimize performance for increased throughput and lower latency. Power efficiency is also a crucial aspect of ML workloads. Observability services span from individual ML workload management to comprehensive datacenter and fleet management.

Before his current role, Aspi served as the Engineering Director for GCP Cloud Networking at Google. In this position, he led teams that built the virtual networking infrastructure and all essential networking functions from the ground up, making GCP a viable platform for enterprise businesses. Google Cloud is recognized as a vital and expanding segment of Google's operations.

Prior to his work on Cloud Networking, Aspi was responsible for teams that developed front-end serving systems. These systems provided a virtual global computing platform, offering unified query distribution, serving, load balancing, and DDoS protection. Aspi brings extensive experience in leading engineering teams to develop distributed infrastructure software systems, with a strong emphasis on reliability, scalability, and optimal performance.

Languages: English, Gujarati

EDUCATION:
School: University of Louisiana at Lafayette
Degree: Master's degree, Computer Science
Activities: None

School: Indian Institute of Technology, Bombay
Degree: B Tech, EE
Activities: None

Number of Followers:
1732