AI Quote Generation in Minutes

Yes. The customer leaves a voice note describing the job — for example, 'I need the bathroom tiles replaced, about 40 square feet, existing tiles need to be stripped' — Whisper transcribes it, the AI parses scope and applies your rate card, and the PDF is sent within minutes. For anything outside the model's confidence — structural unknowns, permit conditions, unusual materials — the estimate includes a flagged line item explaining what requires a site look.

It uses only your rate card. You define labor rates by trade and complexity level, material costs by unit, and any regional adjustments. The model applies those values to the parsed scope. If an item falls outside your rate card, it is flagged rather than estimated. You can update the rate card at any time and it takes effect immediately on the next estimate.

The system generates a partial estimate covering the items it can price with confidence, and flags the uncertain elements explicitly — for example, 'structural assessment required before pricing beam removal'. The customer receives a professional document that shows you have understood their request and gives a clear next step, rather than silence. This keeps the conversation moving without over-committing.

Each generated estimate fires a webhook to your CRM: lead created, scope captured, estimate value attached. If the customer clicks the 'book a visit' button in the PDF, the appointment drops into your scheduling system. We support standard CRM platforms and can connect to custom systems via webhook. No manual data entry.

Yes. Customer descriptions and contact data are processed with a lawful basis, stored for the minimum time necessary and never used to train public AI models. We sign a data processing agreement with every client. CCPA opt-out is supported out of the box, and we configure data retention to match your state requirements.

At launch we configure the trades you define: remodeling, plumbing, electrical, HVAC, roofing, windows, flooring and landscaping are the most common. For each trade we map the scope types — installation, repair, replacement, emergency — and the variables that affect price: area, material grade, access complexity, permit requirement. Unusual or bespoke scopes are flagged rather than priced.

Yes. Whisper supports multiple languages including Spanish. If your firm serves Spanish-speaking customers, the system can accept voice notes and descriptions in Spanish, parse them and generate the estimate in English (or Spanish, your choice). We configure the language behavior during onboarding.

We track three metrics: estimate-to-booking rate (how many estimates convert to a booked visit or signed job), time-to-first-reply (how quickly the customer receives a real number after contacting you) and average job value of AI-quoted jobs versus manually quoted ones. You get a monthly dashboard with those figures and a review call to discuss any gaps.

Neither. It handles the first step — giving the customer a credible, structured estimate fast enough to keep them in the conversation. Jobs that require a site visit still get one: the AI flags them. What it replaces is the 24-72 hour gap between first contact and first number, which is where most construction firms lose the job. Your estimator spends their time on complex jobs that need judgment, not on data entry for straightforward requests.