Construction document control software that reads the set, not just stores it

The expensive failures in document control are not lost files. They are work built to a superseded sheet. A crew tiles to Rev B because Rev C never made it past someone’s inbox, and the rework line item is yours. The drawing existed. The revision was issued. The transmittal even went out. The current version just was not the one the field had open. Document control is the discipline of closing that gap, and most software treats it as a filing problem when it is really a knowing problem.

Storage and control are different jobs. A shared drive can tell you a file exists. It cannot tell you that the file is superseded, that a change order overrides the detail on it, or that the transmittal recorded delivery to the wrong trade. That gap is not free. Inadequate interoperability across the documents and systems on a US facilities project was put at roughly $15.8 billion a year, about two thirds of it landing on owners, and most of that loss is people rekeying, re-checking, and rebuilding context that the documents already held.^[1] Brad reads the set so the current version is the one that answers, not the one that happened to be open.

Control means the current version, every time

Brad tracks revisions, approvals, and issue dates across your set, so the version in force is the one the field actually gets. Ask “what’s the approved spec for the lobby tile?” and the answer comes back with the revision, the approval date, and a link to the document it came from. Not “I think it’s the latest, let me check.” The model holds the revision history, so superseded sheets are marked as superseded rather than sitting one folder over, indistinguishable from the live one.

This is the failure mode that turns into rework, and rework is not a rounding error. Direct rework averages around 5% of total construction cost, and a large share of it traces to bad data: building the wrong thing because the right document was not the one in hand.^[2] A separate FMI and Autodesk study put the rework caused specifically by bad data at roughly $88.7 billion in 2020, about 14% of all rework that year.^[3] Keeping the current version findable is the cheapest rework prevention there is, because the rework never starts.

It reads the documents, it does not just shelve them

A drive holds the file. Brad reads what is inside: plans, specs, submittals, transmittals, change orders, and pulls out the things document control actually runs on. Revision numbers. Approval status and dates. The CSI MasterFormat spec section a sheet carries. Which document governs when two of them claim the same ground. A transmittal stops being a PDF in a folder and becomes a record of what was issued, to whom, on what date, at what revision.

That is the line between storage and control. Storage tells you a document exists. Brad knows which document is in force, and can show its receipts. When a submittal is approved, the approval is attached to the submittal log entry, not buried in an email three reviewers deep. When a transmittal records delivery, the revision it delivered is the one the recipient sees as current, so the next question, “did the framer get Rev C?”, has an answer instead of a shrug.

Catching conflicts and superseded documents before the field does

Because Brad connects related documents, it can surface the quiet conflicts: a change order that supersedes a spec section, two documents that flatly disagree about the same detail, a transmittal that issued Rev B to a trade that needed Rev C. These are the discrepancies that turn into rework two months later, when nobody remembers the conflict was ever there. Brad shows them while there is still time to act.

It does not paper over a genuine conflict with a confident guess. When the drawings and the spec disagree, Brad flags it and points at both sources, because a discrepancy between two governing documents is a human’s call to make, often the architect’s. The value is not that Brad decides; it is that the conflict surfaces during review instead of during demolition. Those unresolved discrepancies are also where disputes start. The most common cause of construction disputes is a party failing to understand or comply with its contract obligations,^[4] and a contract is only as clear as the document set that carries it. The contract itself defines the formal RFI, submittal, and change processes the document set has to keep straight.^[5]

Distribution without the version-control headache

Document control breaks down at distribution. The current set lives in one place; the version on the framer’s phone is whatever PDF reached them last. So instead of chasing the latest file around an email chain, everyone asks Brad and gets the same current answer, with its source attached, over the email and text threads the team already uses. There is no new portal to log into, and no rule that only works when the busiest person on the job remembers to enforce it on a Monday.

This matters because the searching is already a measurable cost. Construction professionals lose hours every week looking for project data and reworking what was built wrong, time one study valued at roughly $177.5 billion a year across the US industry.^[6] Document control that answers “which version is current?” on demand is a chunk of that time the schedule gets back. The transmittal log, the submittal status, the revision history: each one assembles itself as the work happens, rather than getting reconstructed by hand the week before closeout.

Where Brad governs, and where it does not

Brad is document intelligence for construction, not a guarantee or a certification of compliance. It reads your set, tracks revisions and approvals, flags superseded versions and conflicts, and answers with the source attached. What it is not: a replacement for your contract’s formal review and approval process, a stand-in for the design authority that decides which document governs, or a warranty that the set is complete. A person still owns approvals and the answer of record. Your project’s content belongs to you, and each workspace is walled off from every other one. If your organization needs specific data-handling documentation before adopting Brad, ask us and we will tell you plainly what we can and cannot provide.

Document control is one of those disciplines that is invisible when it works and very expensive when it does not. The point is not a tidier filing cabinet. It is that the current version reaches the field before the wrong one gets built, that a superseded sheet announces itself, and that a transmittal can prove what was issued and when. Brad reads the set so that knowing which version is in force stops being a discipline you enforce, and becomes a question anyone can ask.