Liquid Cooling Technicians for AI Data Centers
Installers and commissioning techs for direct-to-chip, rear-door, and immersion cooling.
AI workloads have pushed rack densities past the point where air alone can cool them. Cortex Construct supplies liquid cooling technicians who install and commission the direct-to-chip (DLC), rear-door heat exchanger (RDHx), and immersion cooling systems that serve NVIDIA H100/H200/GB200 class deployments. Our technicians work across the full liquid cooling stack: facility water loop, CDUs (coolant distribution units), manifolds, in-rack quick-disconnect plumbing, and dielectric fluid handling for immersion systems. This is one of the fastest-emerging specialisations in the industry and we are actively building bench in every major market.
Demand Drivers
Liquid cooling is moving from niche to mainstream. NVIDIA reference designs for GB200 NVL72 require liquid cooling at the rack. Every AI-focused build from 2025 onwards includes significant DLC or immersion scope, and the trained technician pool is essentially still being built.
Key Skills
Typical Tasks on Data Center Projects
- Install facility water piping to CDU rows
- Commission CDUs and prove flow, pressure, and delta-T
- Install and leak-test in-rack manifolds and drip trays
- Fill and commission immersion tanks with dielectric fluid
- Support rack-level integration with server OEMs
- Handover water chemistry and leak-detection systems to operations
Required Certifications
Key Data Center Terminology
Specialist terms this trade works with every day. Click any term for a full definition.
A cooling method that uses liquid (water, dielectric fluid, or refrigerant) in direct contact with or very close to IT components to remove heat more efficiently than air cooling. Types include direct-to-chip (cold plates on processors) and immersion (servers submerged in dielectric fluid). Liquid cooling is increasingly required for AI/HPC facilities and demands new plumbing and piping skills from the construction workforce.
A liquid cooling method where cold plates are mounted directly on heat-generating components (CPUs, GPUs) and coolant circulates through them to remove heat. Direct-to-chip cooling is the primary cooling approach for high-density AI server deployments. Construction requires plumbers and pipefitters to install manifolds, CDUs, and distribution piping to each rack.
A cooling technique where IT equipment is submerged in a non-conductive dielectric liquid that absorbs heat directly from components. Single-phase immersion keeps the liquid below boiling point; two-phase immersion allows the liquid to boil and condense. Immersion cooling is used for the highest-density AI workloads. Construction requires specialized tank installation, plumbing, and heat exchangers.
A device that conditions and distributes liquid coolant from the facility's chilled water system to the liquid cooling loops serving individual IT racks. CDUs manage temperature, pressure, and flow of the coolant. Installation involves pipefitters for water-side connections and electricians for power and controls.
Related Guides & Articles
Other Trades We Provide
Contact us
Tell us about your project and we will build a custom staffing plan.