Hemant Sethi

About Hemant Sethi

I build high-performance data platforms and AI-optimized storage systems. Currently at Crusoe, I\'m working on S3-compatible object storage with node-local caching designed for GPU-intensive ML workloads.Before this, I led infrastructure work at AWS on Amazon Firehose, where we processed billions of events daily for enterprise customers. At WeWork, I rebuilt the spatial resource service that powered space management and inventory systems, handling bookings across 800+ locations worldwide.I\'m particularly interested in distributed systems architecture and draw a lot of inspiration from papers like Dynamo, Spanner, GFS, and Raft. I\'m learning more in the areas of RDMA-enabled Storage, Node local & caching strategies for GPU clusters, and storage design for large-scale model training. Always happy to connect with others working on AI infrastructure challenges or exploring research in these areas.

Experience

1. Staff Software Engineer

   Crusoe

   Jul 2025 — Present · San Francisco, CA, US

   Building AI Infrastructure and High Performance Storage Layer.

Education

• National Institute of Technology Jalandhar

  Bachelor of Technology (B.Tech.), Computer Science

  2008 — 2012

Skills

• Gemfire
• Representational State Transfer (Rest)
• Software Development Life Cycle (Sdlc)
• Maven
• Spring Boot
• Eclipse
• Pivotal Cloud Foundary
• Linux
• Javascript
• Jsp
• Spring Mvc
• Jdbc
• Java Message Service (Jms)
• Data Structures
• Tomcat
• Programming
• Java Enterprise Edition
• Mongodb
• Web Applications
• Core Java
• Servlets
• Algorithms
• Jenkins
• Multithreading
• Java
• Xml
• Web Services