Brad for architects: fewer answerable RFIs, propagation tracked, a record that holds up

You stamp the drawing set, and then the real work of defending it starts. The same field question lands three times. A substitution gets approved under schedule pressure and quietly walks away from what you drew. You date-check sheets like a coroner to find the one that still governs. Construction administration is mostly a memory problem wearing a design hat, and the memory is scattered across the set, a prior ASI, and a dozen email threads. Brad’s job on an architect’s desk is narrow and useful: answer more of the answerable questions before they reach you, track where a change has to propagate, and keep a record that survives the eight-months-later phone call.

Start with volume, because the volume is the part that wears people down. One PlanGrid analysis put RFIs near 800 on a typical project, roughly ten for every million dollars of contract value, with an average response time around eight days in the US and more than a fifth never answered at all.^[1] A broader study across 1,362 projects counted something like 1.08 million RFIs in total.^[2] A real share of those are genuine conflicts that need a designer’s call. The rest are the field asking about a flashing detail or a dimension the documents already cover, because writing an RFI is faster than digging through six PDFs to find it. That second pile is the one you want off your desk.

Fewer RFIs reach you, because more of them are already answered

Brad reads the project the way a construction administrator would: the drawings, the specifications, the change orders, and the clarifications and email threads wrapped around them. When a field question comes in, it pulls the governing CSI MasterFormat spec section, the current detail at its latest revision, the right sheet, and any prior ASI or RFI that already grazed the same issue, and lays them next to the question. If the documents answer it, the field gets the answer with the source attached, and you never see the RFI. The third person this month asking about the same parapet flashing gets the same answer you already gave the first two, cited.

That is the difference between an architect’s hour spent on a question that needs an architect and an hour spent re-explaining a detail that is drawn plainly on sheet A-502. Waiting on information the project already holds is non-value-adding work in the lean sense, the kind of waiting that adds days to a cycle without adding a decision.^[3] Brad does not decide the hard ones for you. It clears the answerable ones out of the way so the hard ones are most of what reaches your screen.

When a change lands, Brad tracks where it has to propagate

A design decision rarely lives on one sheet. Change the curtain wall mullion and you have touched the elevation, the wall section, a detail or two, the glazing spec in Division 08, maybe the schedule of values line that prices it. The failure mode in CA is not making the change. It is missing the third place the change had to land. A proposed “equal” substitution gets approved against your 09 21 16 gypsum assembly, and the place it conflicts is a detail nobody reopened. The drift shows up on a site walk instead of on a screen.

Because Brad holds the documents as a connected record rather than a folder of PDFs, a change ties back to the design intent and every document it touches. A superseded sheet still floating in the trailer gets flagged before a crew builds to the wrong rev, which is exactly how a wall gets framed twice. A substitution that walks away from what you specified shows up as a flag while it is still a question, not after it is hanging on the wall. The reason this matters most early is structural: your ability to influence an outcome is highest at design and falls as the cost of changing it climbs through construction, so catching propagation early is the whole game.^[4]

A record of intent that holds up when someone questions it

Eight months after a detail gets built, somebody asks why it was built that way, or whether the added cost was ever authorized. If your record of intent is scattered across inboxes, you reconstruct it under deadline and hope you found everything. If it is connected, the basis is already sitting together: the drawing, the spec section, the clarification, the approval, and the conversation that settled it. Brad keeps that thread tied while the work happens, so the record assembles itself instead of getting excavated at closeout when the as-builts are due.

This is not only a convenience question. The most common cause of construction disputes is a party failing to understand or comply with its contract obligations, and disputes are not small: the average North American dispute reached tens of millions of dollars and dragged on for over a year.^[5] A sourced, dated trail of intent, what governed, when it changed, and who signed the clarification, is the difference between a defensible timeline and a room full of people fairly sure it was around April. The same connected record carries the handoff: project teams turn over, the record does not, so the design intent outlives the people who set it.

Where the architect’s judgment stays the architect’s

Brad is document intelligence pointed at construction records. It does not exercise professional judgment, it does not stamp or seal drawings, and it does not stand in for your RFI, submittal, or formal clarification processes, which the contract defines and Brad works inside of rather than around.^[6] It reads, connects, and drafts from your project’s documents and messages, and it answers with the source attached. A person owns every answer that goes out, and your seal stays on no drawing Brad touched. Your project’s content stays yours, and each workspace is walled off from every other. If you have requirements about how long your record of intent is retained, or how your CA documentation is handled, ask us and we will walk you through exactly how it works.

You will not stop the field from asking, and you would not want to. What you can change is how many of those questions reach you already answered, whether a change is caught while it is cheap to fix, and whether your record of intent is something you can stand behind or something you have to rebuild. The drawing set is where you put the intent. Brad is how it stays whole and one question away for the length of the job.