All articles

Estimating and measurement

Working with AI as a construction estimator

Share the preparation work and keep track of the decisions in your estimate.

In this article

Hundreds of pages of specifications, dozens of construction drawings and a deadline drawing closer. It is a familiar situation for many estimators. Long before the first estimate items are written, considerable time has often gone into reviewing the project information. AI can speed up much of that preparation. But how do you stay in control of the estimate?

Less searching, more time to estimate

Preparing an estimate requires a precise understanding of what will be built. Which materials are specified, what work is involved and which quantities must be included? That information is spread across drawings, specifications and technical descriptions. Sometimes it appears in several places, and the details do not always agree.

AI can help review these documents and bring relevant information together. This includes identifying work, collecting available quantities and preparing estimate items. Using the construction company’s own work descriptions and estimating structure can also help an initial estimate fit its established way of working.

The main saving is in preparation time. Instead of manually searching for and copying information, the estimator can focus on whether it is correct and based on the right assumptions.

What if the specification and drawing disagree?

Take the refurbishment of a commercial building. The technical description specifies a 100 millimetre internal wall. The corresponding detail drawing shows 125 millimetres. It is a relatively small difference, but it can affect the choice of materials and the final cost.

Normally, the estimator has to spot this discrepancy. AI can compare the two documents and draw attention to the contradiction. It can then ask which wall thickness to use, or whether a tender clarification question should be raised first.

The estimator decides what to do next. Other project information may already establish the correct dimensions. If not, the discrepancy can be recorded as an issue to resolve, with a reasoned assumption used in the meantime.

Situations like this show why AI can be useful not only for collecting information, but also for asking questions. Estimators have expertise that is not always recorded in the project documents. Combining that knowledge with AI analysis can help produce a better-supported estimate sooner.

Working together on a substantiated estimate.

A good estimate must remain verifiable

There is, of course, another side to using AI. It can misinterpret information, miss work items or make assumptions that do not fit the project. An estimate can look complete at first glance while still missing important elements.

As an estimator, you therefore need to trace the basis of each item. Where did a particular quantity come from? Was it taken from a drawing or calculated from an assumption? And were all relevant project documents included?

That information allows you to assess the result and correct it where needed. Responsibility for the estimate remains with the estimator.

100 or 125 millimetres?

Technical specification
Internal wall: 100 mm.
Detail drawing
Internal wall: 125 mm.
Preparation with AI
Bring the two source references and the discrepancy together.
The estimator’s decision
Check revisions and the current agreement. Ask for clarification if the documents do not resolve the issue.

Clearly record any provisional assumption. Finalise the material choice and price only once the basis is confirmed.

Our approach at Baudex

At Baudex, we have designed the software so that AI supports estimators in analysing project documentation and preparing estimates. It can use the company’s own estimating structure and work descriptions.

When information is missing or documents contradict each other, Baudex can flag this and ask for further input. The project information used remains traceable, so the estimator can check what the results are based on.

The principle is simple. Let AI handle preparation where it can, and use the estimator’s knowledge and experience for the decisions that matter.

Want to see how this works in practice? Discover how Baudex can support your estimating process.

Explore draft estimating

Choose language

Dutch is selected.