Energy recovery for AI infrastructure

AI is an energy problem.
We're engineering the recovery.

EnGen is developing energy recovery systems that capture waste heat from AI data centres and convert it into useful energy — designed to work alongside the infrastructure already in place.

Get in touch See how it works

🖥 AI Compute

♨ Heat rejected

⊕ EnGen

⚡ Power recovered

The Energy LoopClosed

The problem

A structural gap
in every AI data centre.

~1.5%

of global electricity consumption goes to data centres — and it's growing fast.

Demand is projected to nearly double by 2030 as AI workloads scale.

IEA, Energy and AI (2025)

Nearly all of the electricity an AI data centre consumes is ultimately converted into heat. Most of that heat is currently rejected rather than reused.

AI infrastructure is designed to run at full power — not to reuse what it wastes. Every rack of AI compute rejects a continuous stream of heat that represents recoverable energy.

The technology to capture low-grade heat has existed in research settings for years. What's been missing is a system engineered specifically for the thermal profile of AI infrastructure — distributed, continuous, and increasingly liquid-cooled.

AI infrastructure is designed to reject heat — not reuse it.

The solution

A closed loop,
built into AI infrastructure.

Designed for AI data centre infrastructure.

EnGen sits in-line with the liquid cooling loop of an AI data centre, recovering energy from coolant that's already carrying heat away from the racks — without disrupting the loop it depends on.

Two outputs, not one: a portion of recovered energy is converted to electricity, the rest is delivered as useful heat.

Heat is generated

AI server racks generate continuous low-grade heat through liquid cooling.

Warm coolant exits the racks

Warm coolant exits the racks carrying recoverable thermal energy.

EnGen sits in-line

The EnGen Energy Recovery Module sits in-line with the liquid cooling loop.

Converted to electricity

A portion of the recovered energy is converted into electrical output for facility use.

Useful heat, routed onward

Remaining thermal energy can be routed to heating, storage, or nearby heat networks.

The loop closes

Cooled coolant returns to the cooling loop, reducing wasted thermal energy.

System integration 01 / 06

Warm coolant

Electrical output

Thermal output

Cooled return

Why now

AI infrastructure is
changing.

The rapid growth of AI is fundamentally changing how data centres are built. Higher compute densities and the transition to liquid cooling are creating continuous streams of recoverable thermal energy.

AI Compute

Higher Rack Density

Liquid Cooling

Continuous Warm Coolant

EnGen Opportunity

Growing AI demand

AI workloads continue increasing data-centre power consumption.

Liquid cooling adoption

High-density AI infrastructure is increasingly moving towards liquid cooling.

Recoverable energy

Continuous warm coolant creates an opportunity for practical energy recovery.

Get in touch

Interested
in EnGen?

We're building something real.
We'd love to hear from you.

Whether you're a data centre operator, an AI infrastructure company, a research collaborator, a strategic partner, or an early-stage investor — reach out directly.

Rebecca Turnock

Founder, EnGen

Hi, I'm Rebecca, the founder of EnGen.

We are currently speaking with AI data centre operators, cooling specialists, researchers and investors as we validate our technology and refine where it fits into future infrastructure.

If you work in this space or you're interested in what we're building - it would be great to have a conversation.

AI is an energy problem.We're engineering the recovery.

A structural gapin every AI data centre.

A closed loop,built into AI infrastructure.