You're Only Capturing 60% of Site Issues | Gather

15 Feb 2026 · 1 min

Watch on YouTube

0:45

In this short discussion, the speakers explore how AI‑driven automation can replace the labor‑intensive task of junior staff manually reviewing construction site records. Currently, quantity surveyors (QS) rely on inexperienced team members to sift through contracts, clauses, and on‑site data to spot potential compensation events and early warnings, a process that only captures about 60% of issues. By deploying an automated records‑agent, the system can flag anomalies and potential compensation events in real time, giving QSs a concise alert rather than a massive data dump. This enables QSs to focus on higher‑value activities—pricing, risk assessment, and program impact—while potentially increasing the detection of compensation events by up to 40%. The conversation emphasizes that the goal is not to replace QSs but to augment their efficiency, improving risk visibility, schedule forecasting, and cost‑to‑complete estimates.

Chapters

How is AI impacting what we're doing? I think we're going to start to automate, and we'll use the word automate rather than AI. We're going to automate laborious processes that we hate doing. We often get our most junior person to go and start to review records, go through all that set of data. We can now automate that process. At the moment, we rely on the most junior person, inexperienced person, to go through site records, expect them to know the contract inside out. Expect them to know how to build a bridge, expect them to know exactly what you signed up to in your clauses. But really, we can do that automatically. So I think we can actually start to take that grunt work away from people and have the data flag what those issues are.
So it's a, it's a records agent, not a kind of AI QS.
We're not trying to replace the QS here, we're trying to speed the QS up so they can get more done probably what you want, but we're basically taking that grunt work away from the QS so they can do the things that add value. How can we make your QS's life easier? Well, we can tell them when something's wrong, when there's a potential compensation event, potential early warning needs to be raised. Rather than having to go search through the data, you now get someone telling you. We're only capturing 60% of the issues, so imagine if you as a QS now have 40% more compensation events or early warnings that you are notified about. Like, you won't raise every single one of those to the customer or the client, but at least you'll be able to price them, understand the impact. So is it your risk? How does it affect your program? How does it affect your cost to complete?