Fleets at scale
I've worked on platforms running hundreds of thousands of machines across dozens of datacenters. At that size something is always broken, so the job is making failure routine: notice it, isolate it, repair it, put it back.
I build the layers between a GPU and the model running on it.
Infrastructure engineer since 2018, across public, private and GPU clouds. Senior Software Engineer at Lambda, previously Tesla. Pick a layer to see where I come in.
What I bring to an infrastructure team
Control planes
Kubernetes operators, controllers and the services around them. Most of what I build is a loop that compares what's running with what should be running and closes the gap.
Public, private and GPU clouds
I've built on AWS, Azure and GCP, helped build a private cloud from the ground up, and now work on a GPU cloud. Having done all three, I know what each is good at and what it costs to move between them.
Observability
Metrics, logs and events wired through every layer, and real-time state instead of stale snapshots. Infrastructure should notice its own problems before the people using it do.
Virtualization and shared GPUs
Hypervisors at scale, GPU VMs, and the placement and quota logic that lets many teams share expensive hardware safely.
Migrations that save money
I've moved tens of thousands of existing machines onto new platforms without losing track of who owns them, and helped move workloads from public cloud onto owned hardware, saving millions of dollars a month.
What colleagues say
“He quickly became a VM as a Service expert within Tesla Cloud project and a versatile contributor across the platform. His readiness to assist and mentor new engineers accelerated their integration and productivity, making him the first point of contact for many. His keen ability to spot and rectify design flaws has significantly influenced feature development.”
Hassan Youssef, Sr. Staff Software Engineer, Tesla
“He picked up the work very quickly and was very thorough in his approach every time we discussed on a problem, be it technical or functional. The best part I liked was his attention to minute details.”
Angshuman Das, Software Engineer, Google
“He is down to earth and always completes tasks before the deadline. He is very good in cloud computing technologies.”
Neela Shiva Prasad, Senior Software Engineer, Google
Where I've worked
Education
- Master of Science, Computer Science and EngineeringSanta Clara University
- Bachelor of Technology, Computer Science and EngineeringJawaharlal Nehru Technological University
A bit about me
I like hard problems, and infrastructure keeps handing them to me. I started out full stack, so I care as much about the API an engineer touches as the server underneath it.
I live in San Francisco. Off the clock you'll find me on long drives with house and techno on, watching football or playing FIFA, and stopping for every dog I pass. I'm fairly sure I speak fluent dog.