Quality decisions, faster: How a Tier 1 global data center developer transforms site selection with Build.
“For a long time I’ve worked on projects for hyperscale, including at AWS. With Build, we can make quality decisions, faster. Build helps us find the fatal flaws in our pipeline, and helps us find and move faster on the sites that can work. Other developers are sleeping on this. I’m happy we’ve partnered with Build since the early days.”
- Mark Gethings, Senior Manager – Site Selection & Development Due Diligence (EMEA)
The Customer
A Tier 1 global data center developer operating across EMEA. Mark Gethings leads site selection and development due diligence for the region, responsible for identifying viable sites and eliminating risk across power, planning, fiber, and environmental constraints before capital is committed. The function requires simultaneous breadth of market coverage and depth of technical validation across multiple European jurisdictions.
About Build
Build is the AI-native operating partner for institutional real estate. It delivers site selection and due diligence end-to-end as an outcome, not software. AI workers execute the full delivery across two services: site sourcing and technical diligence. All findings are scraped and stored in the Vault, Build's proprietary data repository with a built-in citation engine that traces every output back to its source. Scoring is handled by a mapping layer within the evaluation stage, custom-tailored to each client and calibrated continuously across projects. Outputs are delivered as site sourcing presentations built on infrastructure data from over 1,600 sources, sent directly by email so clients receive decision-ready work without leaving their existing environment.
The Problem
Site selection for data center development faces a structural constraint: breadth and depth are in direct competition. Teams can either cover the market at risk of missing critical constraints, or go deep on fewer sites at risk of overlooking better opportunities. Compounding this, the process runs sequentially. Power specialists, planning consultants, and environmental experts operate in distinct phases, creating coordination overhead and timeline drag. Work spanning multiple European jurisdictions routinely took weeks to months before reaching a decision point. These decisions are critical and lead to transactions of hundreds of millions of dollars.
"Before Build, the process was inherently sequential. By the time individual power, planning, and environmental consultants had each provided their early reports, weeks had passed - and sometimes that's when you'd find the fatal flaw. Covering the market at the depth institutional decisions require wasn't something one team could do at scale."
The Mechanism
Build's AI workers run site sourcing and technical diligence as concurrent services. Sites identified across European markets are immediately evaluated across grid capacity, fiber access, planning constraints, environmental risks, physical conditions, and community sentiment, in parallel rather than in sequence. All findings are scraped and stored in the Vault, with Build's citation engine maintaining a traceable link from every output back to its underlying source across more than 1,600 data sources.
Scoring is applied through a mapping layer within the evaluation stage, configured to each client's specific requirements and refined continuously across engagements. The output is a site sourcing presentation: structured, standardized, and delivered directly by email. Clients receive decision-ready work without adopting new tooling or changing how they operate.
"What changes with Build is that you're not waiting for specialists to hand off to each other. The critical constraints across grid, planning, political sentiment and environment come back together, and they come back fast. We receive structured, source-backed outputs directly by email and added to your own personalised dashboard - no new systems, no change to how we work. It slots into our process rather than replacing it."
The Evidence
Earlier risk detection. Constraints in grid, planning, or environment that previously surfaced late in diligence, after time and capital were committed, are identified at screening stage. This shifts risk detection from late-stage diligence to early-stage screening, reducing wasted effort on non-viable sites to give development teams more bandwidth to focus on what matters; insight, judgement and relationships.
"Finding the fatal flaws early is everything in the data centre business. Time spent advancing a site that can't work is time and capital you won't get back while competition moves forward. Build surfaces those constraints at screening stage in a detailed manner, before we've committed resource. That's not a marginal improvement - it changes what the team can focus on and allows a you to be highly prepared at the early engagement piece of development."
Faster stakeholder engagement. Structured, validated outputs enabled Mark's team to approach landowners with full constraint and opportunity clarity before initial conversations, previously not possible within standard diligence timelines. The output is designed for action, not review. Findings are structured so they can be used directly in discussions with stakeholders, including landowners, advisors, and internal decision-makers.
"We can now go into a conversation with a landowner having already validated the grid position, the planning exposure, the environmental constraints, and any existing political backlash. That level of preparation, at that stage of the process, simply wasn't achievable before. It changes the quality of the conversation and how we're perceived as counterparties."
Source-level verification. Every output in the site sourcing presentation traces back to its source through Build's citation engine. Rather than service-level links, each source is stored in Build’s Vault with the exact relevant passage highlighted, whether it appears on page one or page one hundred of a document. For institutional developers making capital commitments, the difference between a link and a verified excerpt is material. Particularly when operating in countries where language can be a barrier.
"When you're making large capital allocation decisions across multiple European jurisdictions, you cannot rely on a link to a document. You need to know exactly what it says and where. Build's citation engine gives you the specific passage, in context and stored in your own personalised dashboard. In countries where language is a barrier, this is essential."
A Structural Shift in Development Workflows
Site selection is no longer a funnel of elimination over time. It becomes a parallelized system where opportunities are identified, validated, and ranked simultaneously.
This changes both speed and outcome quality:
- Faster elimination of non-viable sites
- Earlier identification of high-probability opportunities
- More consistent decision-making across geographies
"Site selection used to feel like a process of weeks of manual work, gradual elimination and slow churn. With Build, it's a parallel system, as you're identifying, validating, and ranking site opportunities at the same time. For a small team covering EMEA at institutional scale, that's a fundamental shift in what's operationally possible."
Looking ahead, both Mark and Build see institutional-grade AI as a catalyst for a new era of development: where institutional projects, for the first time, can move at the speed of technology.