How the civil engineering team's pre-submission check supported review.

Turn past comments into requirements the preparer could use.

The relevant knowledge was spread across senior engineers, county requirements, comment letters, and drawing markups. We worked with the senior engineers to capture those requirements and connect them to their sources, so a flag could show the preparer what needed attention and why.

The check identified likely omissions and inconsistencies against those requirements. Its output pointed to a checklist item, requirement, or previous comment for the engineer to inspect. Questions against the same material helped preparers find the requirements while producing the drawings.

Keep professional review with the licensed engineer.

The readiness check supported the preparer's first pass before the set reached a senior engineer. It did not approve drawings or issue a professional seal. Where a flag could not be grounded in a source, the issue remained for human attention.

Senior engineers helped define the examples used to evaluate the check. We tested its ability to find relevant issues and limit unnecessary flags, then checked it on held-out examples before rollout. These were checks of the supporting tool, not a substitute for professional review.

Put the finding beside the drawing that needed attention.

Our initial separate report created friction during rollout. One team dismissed flags in bulk under deadline pressure, so we worked with the client to move the findings into its existing drawing-review tool. Engineers could then inspect a flag in the place where they were already reviewing the set, and the dismissal rate fell the next week.

Separate the operating measures from the financial follow-up.

The six-week pilot follow-up reported first-submission approval increasing from 29% to 61%, senior review time falling from 14.2 to 6.1 hours per plan set, and review cycles falling from 2.4 to 1.3 per set. The later quarterly check reported 66% first-submission approval. These are the client's reported rates for the measured workflow, not a controlled estimate for every project the firm delivers.

The senior-review reduction is 8.1 hours per set, equivalent to just over one eight-hour workday. Median waiting time for senior review fell from seven to two business days. Review effort and queue time are separate measures, so they are not added into a combined time saving.

Delivery ended in April 2025. At the twelve-month follow-up in April 2026, the client confirmed operating margin on net revenue of 20.4%, compared with 15.2% at baseline. The change is 5.2 percentage points, or approximately 34% more operating profit per dollar of net revenue. It does not establish a 34% increase in total operating profit, which also depends on the revenue base.

The client attributed about 2.5 to 3 percentage points of the margin change to the program and the rest to business mix and market conditions. The reported 54% reduction in unpaid land-development rework hours is an operating result, not a second financial benefit added to margin.

Give the team a way to test its own updates.

We supplied an editor, operating instructions, and training so the client's team could change the requirements and test the effect before release. The team demonstrated that capability by adding another jurisdiction, running the tests, and publishing the update without us operating the system.

That handover mattered because county requirements would continue to change. The team could maintain the check and extend its coverage while keeping review and testing in place.

For an operating or technical leader, the wider case connects this review process with the firm's capacity and margin results. Read the full case study.