Question 1

How does the AI handle handwritten notes on drawings?

Accepted Answer

Province reads handwritten markups — redlines, sketched clouds, margin notes, and field annotations — alongside the printed sheet content. The model is trained on real, marked-up construction documents, so it can distinguish an architect's handwritten revision note from the base drawing and factor it into its analysis. When handwriting is ambiguous or illegible, Province flags it for human review rather than guessing, and the flag links back to the exact sheet and location so an estimator can verify the original.

Question 2

What drawing and document formats can Province ingest?

Accepted Answer

Province ingests the formats preconstruction teams actually work in: vector and scanned PDFs (including image-only scans via OCR), full drawing sets, written specifications, addenda, geotechnical and structural reports, and supporting schedules. There's no need to flatten, rename, or pre-process anything — send the governing set as you received it and Province parses sheets, spec divisions, and revisions in place.

Question 3

Does Province work on early, incomplete sets (SD/DD)?

Accepted Answer

Yes. Province is built for the bid window, so it works on schematic and design-development sets, not just permit-ready documents. On incomplete sets it surfaces missing information, unresolved details, and scope gaps — exactly the items you'd want flagged before pricing — instead of treating absence of information as agreement.

Question 4

How are citations linked to specific spec sections?

Accepted Answer

Every answer and finding Province returns is anchored to its source: the exact sheet number, detail callout, or specification section (e.g. 03 30 00 — Cast-in-Place Concrete) it was drawn from. Province builds an internal index of the set as it reads, mapping each statement to a coordinate location on a sheet or a numbered paragraph in the specs. The citation is rendered as a clickable reference that jumps straight to that location, so a CFO, owner, or tender committee can confirm the claim on the page rather than trusting a summary.

Question 5

How does Province avoid hallucinated answers?

Accepted Answer

Province is constrained to your documents. It does not answer from general training knowledge or invent plausible-sounding details — every response must trace to a citation in the set, and if the documents don't support an answer, Province says so. That citation-or-nothing design is why outputs are verifiable line-by-line and why teams can rely on them for constructability reviews, RFIs, and scopes.

Question 6

How does the platform handle conflicting information across addenda?

Accepted Answer

Province tracks the full lineage of every revision. When an addendum changes a detail, modifies a spec section, or supersedes an earlier instruction, Province compares it against the base set and prior addenda and surfaces the conflict explicitly — showing the original language, the revised language, and which document governs. It applies standard precedence (later addenda supersede earlier ones; specifications and addenda typically govern over drawings) while still flagging the discrepancy for a human decision, with citations to every conflicting source so nothing is silently overwritten.

Question 7

What is the security protocol for uploaded project documents?

Accepted Answer

Project documents are encrypted in transit (TLS) and at rest. Each customer's data is logically isolated, access is restricted to authorized users on your account, and your documents are never used to train shared or third-party models. Province operates on a least-privilege basis internally, maintains audit logging on document access, and supports document retention and deletion on request. For teams with specific compliance requirements (owner data agreements, NDAs, or sector-specific controls), we'll walk through our security posture in detail during evaluation — reach out for our security overview.

Question 8

Are my documents used to train Province's models?

Accepted Answer

No. Your project documents are used only to produce analysis for your account. They are not used to train shared models and are not exposed to other customers. Any model improvement work happens on permissioned or synthetic data, never on your confidential sets without an explicit agreement.

Question 9

Who on my team can access uploaded sets?

Accepted Answer

Access is scoped to your account and the users you authorize. Documents and the analysis Province generates are visible only to members of your workspace, and access is logged. As your team grows we can discuss role-based controls so that sensitive bid documents stay limited to the people who need them.

Question 10

What does Province actually produce?

Accepted Answer

Province returns the deliverables a preconstruction team needs before commitments are locked: citation-backed constructability reviews, severity-ranked RFIs drafted from real conflicts, division-by-division scopes of work, responsibility matrices, and bid-package support. You can also ask the set direct questions and get a sourced answer, the same way you'd ask a senior estimator.

Question 11

How fast is turnaround on a new set?

Accepted Answer

Same day. The first analysis on a new set comes back in a day with no onboarding period and nothing for your estimators to learn — Province works like a fast subcontractor, not another platform to administer.

Question 12

Does Province replace Procore or our existing tools?

Accepted Answer

No — it covers the phase those tools weren't built for. Project management platforms like Procore run the build once a project is underway; Province works the preconstruction window, reading the set to surface conflicts, draft RFIs, and build scopes before a bid goes out. It sits alongside your stack rather than replacing it.

Everything you'd ask
a senior estimator.

The questions teams ask before they trust an AI with the governing documents.

Stop reading the set.
Start asking it.

Everything you'd aska senior estimator.