Book a fit call
The method is the same on every job: the project manual goes in, a cited log and drafted submittal packages come out, and a person reviews before delivery.
Each step keeps the spec page attached to what it produces, so your team checks the work instead of taking it on faith.
Send the project manual as a PDF. SubPro reads it front to back, keeps the page number on everything it finds, and works from nothing but your document.
Each requirement becomes a numbered row: spec section, type, description, and the page it came from. Manufacturers and products fill in where the book names them.
Product data from the manufacturer library is matched to the row, then stacked behind a filled-in cover sheet and transmittal as a single PDF.
An or equal, an owner selection, an unclear note: each one is flagged with the page behind it and waits for your PM. The rest is ready to review.
Two rules hold on every project, and both exist for the reviewer.
Open a row and the spec page opens beside it with the requirement highlighted. A PE checks the row against the paragraph the way you would check a sub's work, and signs off on something they have seen.
The project manual is a working input, not a filing. Once the log, the submittal packages, and the review are delivered, the working copies are deleted. Nothing you send is kept or used to train a model. The data-handling page has the full policy.
Search with the citation attached.
Finding one requirement in a project manual is its own job. Where is the warranty term for the roofing? Which sections call for field mockups? What does Division 01 say about resubmittals? The usual method is keyword search across a dozen PDFs until someone calls the architect.
Because SubPro has already read the whole set, you ask instead. The built-in chat answers from your spec book and your log, and every answer carries the page it stands on. Ask where the extended warranty on door hardware is buried and you get the answer and the Section 08 71 00 page that holds it. Ask which rows are still flagged and the log answers directly.
The chat answers only from your documents. The citation sits beside the answer, so anyone can confirm it at the source. The questions that come up on every job are worked through on the answers page.
Delivery is not the end of a submittal's life. Reviews come back, revisions stack up, transmittals go out. SubPro keeps that history on each row while the job runs: what went out, what came back, and which revision superseded which.
By turnover the record already exists: what was submitted, what the architect returned, and when each item moved. Closeout becomes a check instead of a rebuild. The full lifecycle is on the construction submittal software page.
No. The project manual is a working input. Once the deliverables are out, the working copies are deleted behind them. Storage stays your system's job.
Every row carries the spec page it came from. The reviewer opens that page, reads the paragraph, and confirms the requirement instead of taking the row on faith. Rows that need a decision are flagged with the reason.
An editable submittal log in Excel, the submittal packages as PDFs where product data exists, and a flags sheet. The output page shows each file.
Send one spec book through Build my log. A drafted log comes back for review with no call required, and any pilot conversation starts from real output.
Yes. SubPro drafts and organizes. Your team confirms the requirements, rules on substitutions, and owns what goes to the GC. SubPro does not promise a perfect extraction, which is why every row cites its page.
Run the four steps on your own project manual. The drafted log and a finished submittal package are on the table before any commitment.