Decomposes a project scope description, set of drawings, or spec section into CSI-mapped trade packages and drafts a complete RFQ for each — scope language, included/excluded clauses, qualification questions, response template, and your firm's standard terms. The package the procurement team would have written in 4 hours, ready in 60 seconds.
An RFQ that doesn't include the right pieces gets back ambiguous, non-comparable bids — which means hours of leveling work later. This engine ships an RFQ with all the right pieces by default.
Project name, address, owner, GC contact, schedule milestones the trade touches, site logistics summary, contract type (lump-sum / GMP / cost-plus). Auto-filled from the project record. Gives the vendor enough context to price intelligently.
CSI-mapped scope language for the trade package being bid. Pulls from spec sections (when uploaded), drawings, and your firm's reference scope library. Written in vendor-facing language — not internal jargon.
Explicit list of what the vendor is and isn't responsible for. Closes the "by others" ambiguity that drives 80% of post-award disputes. Per-trade defaults from the firm's library; project-specific edits ride on top.
10–15 questions the vendor must answer: bonding capacity, current backlog, similar-project references, key personnel, anticipated subs (if any), safety record, COI tier. Standardized so leveling is apples-to-apples.
A structured response form so every vendor's bid comes back in the same shape: line-item pricing, schedule, payment terms, retainage, exclusions. Eliminates the "each vendor uses different categories" problem at leveling time.
Your firm's standard subcontract terms attached as the assumed contract base. Vendors who can't accept the terms call out exceptions in their qualifications — you see this BEFORE you award, not after.
The hardest part of writing RFQs is the breakdown — turning "build a 200-unit mid-rise residential building" into the right number of trade packages. Too few packages and you're paying GC markup on everything; too many and the buyout effort doesn't justify the savings. The engine reads the scope and proposes a CSI-level decomposition.
The engine PROPOSES a decomposition; the PM confirms or adjusts. Merge two packages into one. Split a package the other direction. The engine learns your firm's preferences over time — some firms always bid drywall + paint together, others always separate.
Every firm has a scope library — the way YOUR estimators describe a plumbing rough package, the inclusions YOUR firm always insists on, the qualifications questions YOUR firm asks. This engine reads against your library, not someone else's.
The engine extracts your scope patterns, inclusion/exclusion conventions, qualification questions, and response template shape. 10 minutes of work. No template-building.
When your team modifies an RFQ after generation, the engine notices the change and asks if it should update the library. Patterns improve job over job without anyone maintaining a master document.
What your firm includes by default for HVAC isn't what you include for concrete. Per-trade library entries handle the variations. Geography-based variations (different code requirements per state) carry too.
The vendor gets a 3-file package:
Vendors who try to bid "outside the template" (some always do) get a polite reminder from the engine: please use the response template. Significantly reduces leveling pain at award time.
RFQ Generator is the first link in the procurement chain. Its output sets the quality of every downstream engine.
~85% on first-day deployment, ~93% after the engine has learned your firm's preferences (most firms make 2-3 modifications to the proposed decomposition on the first job, then ~0 by job 4). The PM always confirms the breakdown before RFQs generate — nothing ships without review.
Yes. You mark the trades/items the owner is buying direct, and the engine excludes them from the RFQ packages and adjusts the GC's scope language accordingly. Useful on hospitality, lab, and healthcare projects where equipment is often owner-furnished.
Yes — with reduced precision. With just a project description, the engine produces a generic-but-credible RFQ structure. With drawings + specs, the scope language references specific spec sections and detail callouts. Specs make the RFQ ~30% sharper on average.
Works with any contract base. The "standard terms" section attaches whatever document the firm specifies as its base subcontract — AIA A401/A441, ConsensusDocs 750, EJCDC C-700, or a bespoke owner-modified document. The engine reads against your terms, not generic ones.
It's a template, not a hard requirement — vendors who bid outside the template still get accepted (better to get a non-template bid than no bid). But the engine flags non-template bids for the PM and pre-leveling work to normalize the format. Most vendors comply after one project of using the template.
Bring a real project scope (or upload the specs/drawings) to a 15-minute call. We'll generate the trade-package decomposition and the RFQ documents live on screen. You keep the output and the firm-library setup.
Start free — 14 days, no card →