Consulting
For hyperscalers and neoclouds.
Elyas Jazayeri, Ph.D. · AI data center infrastructure leader
For hyperscalers and neoclouds.
For investors and developers.
For conferences and panels.
For engineers and program managers.
For a move into AI infrastructure.
Short, practical notes on building the infrastructure behind AI.
Construction takes 18 to 30 months. Getting power to the site can take years longer. Plan the data center around the interconnection, not the other way around.
Racks that used to draw 5 to 15 kW now draw 30 to 80, and the densest systems go higher. The shell can often stay. Almost everything inside it changes.
Main transformers, gas-insulated switchgear, and large breakers can sit in factory queues for 18 to 24 months. Procurement has to start before design is finished.
Price matters less than most teams think. Field execution, safety culture, testing discipline, and the ability to scale decide whether a program delivers.
Journal of Construction Engineering and Management
Journal of Construction Engineering and Management
Journal of Safety Engineering
Journal of Safety Engineering
Construction Research Congress 2018, ASCE
Construction Research Congress 2018, ASCE
Conference paper
Ph.D. dissertation, University of Kentucky
I build the physical side of AI. At Amazon Web Services I lead data center infrastructure programs across North America, delivering the power, cooling, and capacity that AI runs on. I came to it from construction engineering, with a Ph.D. in civil engineering from the University of Kentucky.
My work spans technical program management, MEP execution, high-voltage power, liquid cooling, GPU rack retrofits, utility interconnection, and contractor strategy across more than 100 concurrent projects.
100+
concurrent projects
All
North America markets
Ph.D.
in civil engineering
me@jazayeri.us
Start a request
A few quick questions, then it goes to Elyas