Inside the mold remediation estimate calculator

Preliminary numbers drift. The same basement described by two estimators on two different days produces two different figures, because both are doing arithmetic in their head from experience rather than from a rate sheet. Homeowners notice, and the number that came in lowest sets their expectation for every quote that follows.

Inspecta’s remediation estimate widget splits the problem in half: something has to read the job, and something has to price it. Those are different skills, and only one of them should ever be handed to a language model.

The AI reads the homeowner’s answers and photos and produces a description of the physical job — not a price. It estimates the affected square footage, which materials are involved, how many rooms are affected, what level of containment the work implies, whether removal is likely, how much contents handling is involved, how severe it looks, and how confident it is in all of that.

It is explicitly forbidden from producing a dollar amount, and if a currency figure ever appears anywhere in its output, the report is rejected rather than shown. The model characterises the job. It never prices it.

The scope then goes into an ordinary calculator — plain deterministic code, no model involved — which builds the estimate line by line:

• Containment, at standard or extensive level depending on what the scope implies.
• Material removal when demolition is likely, split between ceiling and wall-type materials by area.
• Cleaning and antimicrobial treatment, always, priced by square foot.
• HEPA air filtration, always, priced per room.
• Contents handling when contents are affected.

Two multipliers then apply: one for heavy or widespread contamination, one for jobs spanning more than a single room. The result is floored at your minimum project price. Because the same inputs always run through the same rates, the same job quoted twice returns the same number.

The calculator ships with platform defaults so the widget works the day you install it, and every one of them is yours to change from your pricing settings.

Rates must be zero or above and multipliers must sit between 1 and 5, which stops a typo from quietly publishing a nonsense number to your customers. If your rates get away from you, one reset restores the defaults.

The homeowner is shown a low and a high figure, with the line items behind them. Nobody can price concealed work from photographs, and a single confident number implies a precision that does not exist before anyone opens a wall.

Line items round to the nearest $25 and the totals to the nearest $50, so the estimate reads like a professional range rather than a machine-generated figure to the dollar. The spread is held open deliberately — the top of the range always stays a meaningful distance above the bottom, so a narrow scope can never collapse into what looks like a fixed quote.

Every estimate carries its own limits. It states plainly that it is preliminary, that it is based on your configured rates applied to a scope inferred from the homeowner’s answers and photos, that it is not a binding quote, and that it can change after an onsite evaluation.

Alongside it sits a list of the unknowns that could move the number: concealed areas that have not been opened, whether the moisture source has actually been repaired, how much affected material is really there once work begins, containment requirements, testing results, and the possibility of particulate spread through HVAC.

That paragraph is doing real work. It sets the homeowner’s expectation before you arrive, so the number on your written quote is a refinement of something they already understood rather than a surprise.

Pricing is shown only by the remediation estimate widget. An inspection and testing widget shows no pricing at all, and an on-site evaluation widget stays about the evaluation. You decide which assessments quote by choosing which widgets you publish.