Hi, I’m Swetal Patel.
I spent 12 years running cloud capacity at Google. Now I help companies spending $50K+ a month on AI infrastructure forecast demand, plan capacity and make commitments they can defend to finance. I also host Tech With Swetal, where I cover the latest trends in the industry.
Served Google Cloud’s largest customers, including Anthropic, Meta and Apple.
Background
I was a Sr. Manager of Cloud Capacity at Google. The job was to make sure the infrastructure behind Google Cloud was there when customers needed it, without buying more than the business could justify. I owned a spend budget above $1B, consistently delivered more than $100M a year in OpEx savings, and avoided roughly $500M of capex.
Along the way I planned capacity for some of Google Cloud’s largest customers, including Anthropic, Meta and Apple, where demand moves fast and the commitments are enormous.
How I think about it
When demand goes exponential, it arrives as a shock: usage quadruples overnight. The teams that handle it do two things at once. They execute relentlessly to keep the lights on today, and they build a sensible strategy for the foundation they’ll need tomorrow. AI infrastructure is that situation at scale, with new models, services and hardware arriving faster than contracts can keep up.
Most FinOps tools explain last month’s bill. I focus on the decisions made before it: how much capacity to plan for, which commitments to sign, and how to prove the ROI.
How we can work together
Start with the free FinOps assessment. If it’s a fit, book a free, screened 15-minute discovery call. From there, I offer a fixed-price AI capacity and forecast review, and ongoing advisory to keep the plan current.
Tech With Swetal
On Tech With Swetal, I break down the latest in AI infrastructure: new innovations, what they cost, and how to plan capacity.