28/08/2026
AI in construction: Scaling success starts with strong foundations
The UK construction sector is facing a significant opportunity. The Government’s latest Infrastructure Pipeline contains 734 major projects worth £718bn over the next decade, with an estimated requirement for between 629,000 and 706,000 construction and infrastructure workers every year for the next five years.
For businesses operating across construction and infrastructure, there is potentially a huge amount of work ahead. But there is also a difficult question to answer: can businesses scale quickly enough to take advantage of it?
The challenge is particularly significant given the sector’s recent performance. Construction recorded 3,931 company insolvencies during 2025, the highest total of any sector, while the wider market remains in an uneven recovery. Infrastructure is offering stronger long-term prospects, but residential construction and parts of the private market continue to face pressure.
For construction businesses, preparing for growth isn't simply about winning more work. It is about having the operational foundations in place to manage it.
But AI is only as useful as the information and systems supporting it
There is understandably a lot of interest in what AI could do for construction. From reporting and document management to data analysis and repetitive administration, there are plenty of areas where AI could help businesses reduce manual work and make better use of the information they already have.
But AI is only as useful as the information and systems supporting it. Many construction businesses have developed their technology in much the same way they have developed their operations: piece by piece. Site management might have one system, commercial another, procurement another, with health and safety, compliance and reporting sitting across further applications and spreadsheets.
That can work when a business is smaller and the volume of information is manageable. As the business takes on more projects, more subcontractors and more reporting requirements, the gaps between those systems can become a significant operational problem.
Information gets copied between applications. Reports are assembled manually. Spreadsheets become the connection between systems that cannot communicate with each other. The problem isn't necessarily that a business doesn't have enough software, in many cases, it has too much and the real challenge is making those systems work together.
From fragmented systems to connected operations
Construction businesses are already managing increasingly complex information across multiple projects.
There is project documentation, compliance information, inspections, environmental reporting, supply-chain data and commercial information, all of which needs to be accurate, accessible and available to the right people at the right time. When systems aren't connected, people end up filling the gaps.
Andy Brown, Director and Co-founder of Propel Tech, explains:
"We see businesses with very good software that isn't necessarily giving them the value it could because the systems around it aren't connected. The technology itself isn't necessarily the problem, the lack of connection between it is."
That doesn't necessarily mean starting again. For many construction businesses, there are opportunities to modernise existing systems and connect them with newer applications, allowing information to move more effectively between different parts of the organisation.
This can reduce manual administration, improve visibility and give managers a clearer picture of what is happening across the business. More importantly, it creates a much stronger foundation for automation and AI.
Building the foundations for AI
AI needs reliable information to work with. If data is fragmented across different systems, information is inconsistent, or critical processes still depend on manual workarounds, AI can't solve those underlying problems.
That makes AI adoption as much a software engineering challenge as an artificial intelligence one. Before businesses can start looking at where AI might create value, they need to understand how their existing systems, workflows and data fit together.
That might mean modernising a legacy application rather than replacing it, or integrating systems that have never previously communicated with each other. Or it could mean creating bespoke software that brings information together in a way that supports the way the business actually operates. The important thing is to start with the problem, rather than the technology.
Turning growth into opportunity
The scale of the Government's infrastructure pipeline means construction businesses have a significant opportunity ahead. Major schemes such as the £10.2bn Lower Thames Crossing are moving into delivery, while planning permission has been granted for the £5bn Universal UK resort in Bedfordshire.
But growth at this scale will put further pressure on businesses to manage more information, more projects and more complex requirements without simply adding more administration, and that is where better-connected technology can make a difference.
At Propel Tech, we work with organisations across the property, construction and facilities sectors to modernise systems, integrate applications and build bespoke software around the way businesses actually operate.
Rather than assuming businesses need to replace everything they already have, we look at what is working, where the gaps are and how existing technology can be made more effective.
Once those foundations are in place, businesses can start identifying where automation and AI can add genuine value, whether that's in reporting, document management, data analysis or repetitive administration.
The principle is straightforward: construction projects bring many different disciplines together, so the technology running the business should do the same. For construction businesses preparing to scale, the opportunity isn't just about taking on more work. It's about making sure the systems underneath the business are ready to support it.
You may also be interested in:
Is operational complexity quietly slowing your business down?
Why AI is making businesses rethink their software foundations
Why “buy or build?” is the wrong question of AI-era software