Carlos is an entrepreneur, researcher and all round life-long learner who enjoys challenges. He studied AI and computer science (BSc, MSc and PhD). Professionally he's always gravitated towards building software tools to make new technology more accessible. With experience in mobile gaming, AR/VR and an startup exit under his belt, he now focuses on making AI compute more affordable by turning wasted GPU capacity in data centres into useful fuel for AI workloads.
Company bio
While hyperscalers invest billions in new data centers, up to $49B in idle compute sits untapped across existing data centers.
We’re building a new kind of hyperscaler — one that doesn’t own or borrow GPUs, but aggregates and programmatically clears global fragmented capacity.
Kalavai turns this underutilized capacity into production-ready AI infrastructure at near-zero cost. Their platform helps AI teams cut cloud spend by up to 90% and scale 3x faster from prototype to production, while data centers double their utilization and revenue using the hardware they already own.
Led by PhD in Distributed Systems and ex-Googler (GTM StratOps, scaled $1B SaaS business)
Backed by Plug & Play, ForumVC and MBA Ventures