Start with the workflow, not the model

Most AI projects in construction start with the wrong question: “which tool should we use?” The better question is “which workflow costs us the most time and error, and would automation actually help?”

On real projects the answer is usually unglamorous — RFI and ITP tracking, re-keying the same numbers across systems, assembling handover documentation. These are high-volume, rule-heavy, and painfully manual. That is exactly where automation earns its keep.

So pick one such workflow. Map what “better” looks like in numbers. Build against the real tools and data, in small increments you can see working. Measure the time saved, then expand. The model matters far less than choosing the right thing to point it at.