In a traditional design-bid-build project, drawings are 100% complete before the GC prices anything. In design-build, you're pricing from 30% documents. Sometimes 15%. You win the work, then finish designing it — and every assumption you made at bid stage becomes either margin or a dispute.
That's a different kind of risk. And it's one that generic AI tools, built for neither construction nor design-build delivery, aren't equipped to handle.
Design-build's growth is accelerating. The DBIA's 2026 conference put pre-construction AI for design-build delivery on its main agenda for the first time. The industry is paying attention. The question is whether the tools are ready.
Scope gaps in design-build don't happen randomly. They follow a pattern — and that pattern starts in pre-construction.
On a DB pursuit, architectural drawings may be at 30% while structural is at 20% and MEP hasn't started. Sub scopes drafted from these documents capture what's shown, not what's implied. When documents evolve, the original scope sheet doesn't always follow.
A $45,000 stone-depth mismatch between civil/structural and architectural drawings on a single slab — caught in a Provision scope review — shows how quickly drawing conflicts add up. On a DB project, those conflicts multiply because the design is still moving.
Most DB contracts contain language requiring the GC to include work that is "readily inferable" even if not explicitly shown. Owners and their counsel know this language well. Subs know it too — and they're increasingly quick to use it.
As one Estimating Manager at a Canadian ICI GC put it: "We have less subs who just kind of a gentleman's agreement… they've become more quick to clarify that we're not including that one piece of scope."
When your scope packages don't address inferred work explicitly, every ambiguous line item becomes a negotiation at buyout — or a change order in the field.
The Scope Gap Playbook — built from 200+ GC interviews — identifies the habits that make scope gaps worse on every project type. On DB projects, these anti-patterns are especially damaging:
For a deeper look at how these habits compound across bid, buyout, and field, see Chapter 4: Anti-Patterns in Scope Language.
The numbers are not hypothetical. The Arcadis 2025 Global Construction Disputes Report puts the average U.S. construction dispute at $60.1 million. "Errors and omissions in contract documents" has been the top dispute cause for six of the last nine years.
FMI's Construction Disconnected report estimates $31 billion in annual U.S. rework costs from miscommunication and bad project data. Twenty-six percent of that rework traces back to communication breakdowns. Twenty-two percent to bad project data.
In design-build, both causes are active at the same time.
Operator-cited examples from the Scope Gap Playbook show what individual gaps look like at the project level:
None of these gaps were caught before bid day. On a design-build project — where the GC owns the design risk — every one of them would have been harder to recover.
Provision is built specifically for general contractor pre-construction workflows. It reads your full project set — drawings, specifications, contracts, addenda — and produces structured, bid-ready outputs. It is not a generic AI tool adapted for construction. It was designed for it from the ground up by a civil engineer and a quantity surveyor.
Scope Agent takes your design-build project documents and generates a complete scope-of-work package — trade by trade — in under 60 minutes. That replaces 30 to 40 hours of manual estimator review per bid.
For DB projects, this matters for two reasons:
Scope Agent surfaces trade conflicts, flags missing callouts, and identifies items that are implied by the design but not yet specified. It doesn't just read what's there. It identifies what should be there.
Design-build contracts carry more risk exposure than most. Owner-favorable language, compressed schedules, and design-liability provisions are standard. Most estimators don't have time to read every clause.
Risk Review runs your DB contract against a pre-built construction risk checklist. It achieves 99.5% accuracy on pre-built checklists and 97%+ on custom ones. It has identified over one million risks across the documents it has reviewed.
For teams managing multiple DB pursuits at once, Risk Review flags high-exposure clauses — scope of design, indemnification, schedule-of-values requirements — before you commit to a price.
DB project sets move fast. Drawings get revised. Specs get issued. Addenda come in the week before submission. Your team needs to find answers quickly — and they need to trust those answers.
Chat Agent lets estimators and pre-construction staff ask plain-language questions against your full project set. "What does Division 07 say about air barriers?" "Does the structural set reference fire-rated framing?" Cited answers in under 20 seconds, with the source document and page number attached.
Across 50,000 queries answered on real construction documents, Chat Agent delivers 95% verified accuracy.
EllisDon — one of Canada's largest general contractors — used Provision on a complex project and identified $1.8 million in scope and risk exposure before bid day. Read the EllisDon case study for the full breakdown.
Provision has reviewed $100 billion in project value across 66,000 processed documents. The accuracy and the speed are not theoretical. They are documented.
Not every AI tool is the same. Here is what to look for when evaluating pre-construction AI for a design-build workflow:
Generic AI tools — including general-purpose tools like ChatGPT — don't read full project sets the way Provision does. They don't produce structured scope packages. They don't flag drawing-to-spec conflicts. And they weren't built by people who have priced a construction project.
If your team is evaluating options, the Provision for general contractors page covers how the platform fits into a full pre-construction workflow.
Design-build is growing. DB contract risk is real. And the document complexity of a DB pursuit — incomplete drawings, evolving specs, owner-favorable contract language — creates scope exposure that doesn't resolve itself at buyout.
As one Pre-Construction Lead at a Top-ENR Canadian GC said: "If you miss anything, they'll bill it."
Pre-construction AI built for GCs — not adapted from a generic tool — gives design-build teams a way to close those gaps before bid day. That's the point of Scope Agent. That's why firms like EllisDon are using it. And that's why the number is $1.8 million saved, not a projection.
To see how Provision works on a real design-build project set, request a demo.
In design-build, GCs price from incomplete documents — often 15–30% design. Every assumption at bid stage becomes a contractual obligation. When drawings evolve and scope sheets don't, gaps appear at buyout and in the field. The GC owns those gaps because DB contracts typically include "readily inferable" work language.
Scope Agent reads your full project set — drawings, specs, contracts, and addenda — at whatever stage of completion they're in. It generates trade-by-trade scope packages, flags drawing-to-spec conflicts, and surfaces implied scope. It re-runs in under 60 minutes when your document set is updated.
Yes. Risk Review analyzes DB contracts against pre-built construction risk checklists with 99.5% accuracy. It flags high-exposure clauses including design liability, indemnification, and schedule-of-values requirements — before your team commits to a price.
Provision delivers 95% verified accuracy across real project documents. It has reviewed $100 billion in project value and processed 66,000 documents. Risk Review achieves 99.5% accuracy on pre-built checklists and 97%+ on custom ones.
Scope Agent generates a complete scope-of-work package in under 60 minutes. That replaces 30–40 hours of manual estimator review per bid — a meaningful difference when your team is managing multiple DB pursuits simultaneously.
Provision is purpose-built for GC pre-construction workflows. Generic tools like ChatGPT can answer broad questions but don't read full project sets, produce structured scope packages, or flag drawing conflicts. Provision was designed from the ground up by a civil engineer and a quantity surveyor for construction-specific outputs.
The Scope Gap Playbook covers the habits, anti-patterns, and trade-specific gaps most common in GC pre-construction — drawn from 200+ interviews with estimating and pre-construction teams. For subcontract language specifically, see Chapter 4: Anti-Patterns in Scope Language.
Scope Agent reviews your full project set and generates trade-by-trade scope packages in under 60 minutes.
See Scope Agent