An estimator once told me the scariest sentence in preconstruction is "I assumed the other sub had that." It has the same energy as "someone must have turned off the water." Everyone is confident, nobody is holding the wrench.

Today's news is about a tool built to find the person who assumed, before the bid is due.

How do you catch scope gaps before they cost you? Check the documents against each other first, then check the bids against the combined documents. Most gaps start as a conflict between the specification, the drawings and the model, and a sub prices whichever one it happened to read. AI is useful here because it can compare every source in a tender without getting tired, as long as every flag points back to the source and an estimator makes the call.

What did Volve just add?

According to AEC Magazine, tendering and pre-construction platform Volve has added an orchestration layer that automates more complex checks across specifications, drawings and BIM models. Users can give it a task in plain language, such as checking quantities against the IFC model. The system plans the steps needed, runs the checks across a tender and verifies the results before reporting them, with the source behind each finding. CEO Herman B. Smith named quantities against drawings and specifications as one of the first use cases.

Two details matter more than the headline. First, the checks can run while the estimate is still being developed, so inconsistencies can surface before pricing is finished. Second, each finding carries its source. That second point is what separates a useful flag from a rumor. An estimator can open the spec section and the drawing sheet and settle it in a minute.

For context, Volve raised $3 million in seed funding in August 2026 to expand across Europe, per tech.eu. Its platform has reportedly been used on more than 2,500 projects, mainly by Nordic firms. The AEC Magazine piece, not that background, is what grounds today's point: the manual comparison work is the target.

How does AI find gaps once the sub bids arrive?

The second half of the job happens at bid leveling. Vendor explainers describe the logic simply: combine the scope that all bids carry, compare it against the specification requirements, and flag items that no bidder has included. Those are your orphan scopes, the work everybody assumed belonged to someone else. One vendor example points to an electrical subcontractor that lost a dispute over excavation work after addenda assigned it, which is exactly the kind of thing a spreadsheet misses at 11 p.m. on bid day.

The two checks work together. Document cross-checking reduces the number of gaps that get created. Bid-versus-scope comparison catches the ones that survive. Running only the second leaves you finding problems after the bids are in and the clock is loud.

Where does a human still have to decide?

A secondhand summary of an Engineering News-Record report on Provision's AI agents notes that ProWest Constructors tested the tools to add bid-preparation capacity while keeping estimators and preconstruction staff responsible for reviewing the generated work. The same summary says human review matters most where drawings, addenda, specifications and local trade practices conflict or change during a bid cycle. That matches common sense: the AI can say two documents disagree, but it cannot know that your electrical sub always carries trenching in this county.

A practical routine: run document cross-checks as the estimate takes shape, send the resulting questions out as RFIs before bid day, then run bid-versus-scope comparison at leveling. Keep the source link on every flag, and log which flags were real. After a few projects you will know which checks earn their keep.

The forward-looking thought is modest. The winning feature is not an AI that finds everything. It is one that shows its work well enough that a tired estimator can trust the ten flags that matter and dismiss the forty that do not.

Frequently asked questions

Can AI really find scope gaps in subcontractor bids?

Yes, as a first pass. Vendors describe tools that compare the combined scope of all bids against the specification and flag items no bidder carries. Estimators still need to review the flags, especially where addenda, drawings and local trade practice conflict.

What is the difference between a bid scope gap and a document discrepancy?

A scope gap is work that no bidder priced. A document discrepancy is a conflict between the specs, drawings or model themselves. Discrepancies often cause gaps, because subs price what they read and each reads a different document.

When in the bid cycle should gap checks run?

As early as possible. Checking documents against each other while the estimate is still being built lets the team raise questions before bids land, instead of discovering the hole during leveling.

Sources