MaraponeAI

Programs, one continuous job

5

Programs, one continuous job

Plans, $69 to $1,199 CAD/month

5

Plans, $69 to $1,199 CAD/month

Unlimited use — no credits, no overage

∞

Unlimited use — no credits, no overage

Programs in this suite

1. Blueprint AuditorTakeoff, code review and tender risk, off the drawings.Included from Field, $149 CAD/month · 7 days free

   Audit tools in one pass

   8

   Audit tools in one pass

   CAD/month, 7 days free

   $149

   CAD/month, 7 days free

   Per-drawing fees or AI credits

   0

   Per-drawing fees or AI credits

   Reading a drawing, not scraping a PDF

   Construction drawings defeat generic document AI. Scale lives in a title block, meaning lives in line weight and hatch, and half the information is in details referenced from somewhere else in the set.

   The pipeline handles the set as a set: it resolves scale, follows detail callouts, and keeps every measurement anchored to the sheet it came from — so a number you don't trust is always one click from the drawing that produced it.

   Eight tools, one pass

   Takeoff, code review, tender-risk scanning, RFI backlog triage, scope-gap finding, change-order risk, deficiency lists and meeting-minute tracking. Each one is a distinct question a project team asks of the same set of documents.

   Running them together is the point. A scope gap is only interesting next to the quantity that should have covered it, and a tender clause only matters when you can see the drawing it applies to.

   Open the tab, the set is already uploading

   It runs in a browser on the MaraponeAI engine — nothing to install, nothing to licence, and the same login opens it on the phone on site. Drawings are the most commercially sensitive documents a contractor holds, so they land in a workspace scoped to that account alone, are never forwarded to a third-party AI vendor, and never train anything.

   Use is unlimited on every plan. No per-drawing charge, no AI credits, no overage — a tool people ration is a tool that stops catching things.

   • 

   • 

   • 

   • 

   • 

   • 

2. SpecCheckerTwo contract documents. One of them is wrong.Included from Crew, $299 CAD/month

   Measurable properties compared

   27

   Measurable properties compared

   Named assemblies, 25 CSI divisions

   39

   Named assemblies, 25 CSI divisions

   Conflict types, including self-contradiction

   11

   Conflict types, including self-contradiction

   Both sides, read into the same shape

   A clause in a manual and a cell in a door schedule are the same fact wearing different clothes. So both are normalised into one structure — a single measurable property, of one named assembly, in one canonical unit, carrying the page or sheet it came from. Everything downstream operates on those, which is what makes the comparison possible at all.

   Where both documents state a value for the same property and the values disagree, that is a finding: both statements, both citations, and the exact wording each document used. Fire-resistance ratings, gypsum thickness and core type, concrete strength, steel grades, R-values and U-values, SHGC, airflow, filter efficiency, sprinkler design density, interrupting ratings, glazing make-up, hardware grade, warranty terms.

   The conflicts a document has with itself

   The obvious check is manual against drawings. Two of the eleven conflict types are stranger and more common than anyone expects: a manual written by three people over two years that contradicts itself, and an old general note on one sheet that the revised note on another sheet was supposed to replace.

   There is no precedence rule to fall back on inside a single document, which is exactly why those have to be asked rather than resolved. It also catches addenda that were never carried through — the two-page PDF that arrives by email during tender and changes one number while the manual and the drawings both keep saying the old thing — and cross-references to sections that were cut.

   Silence looks exactly like agreement

   The audit can only report a conflict where both documents say something about the same property. Where the drawings are silent there is nothing to disagree with, and on a findings list that is indistinguishable from the two documents matching. A tool that hid this would be quietly worse than no tool, because a short list reads as reassurance.

   So coverage is a first-class screen, not a footnote: what was checked, what could not be read confidently, and where the silence is. Findings are then sorted by what stops a material order today rather than by how many there are, and one click drafts the RFI — one letter per spec section, quoting both documents and asking which governs, because deciding that is the consultant of record's job and not the software's.

   • 

   • 

   • 

   • 

   • 

3. AI EstimatorFrom measured quantities to a number you can stand behind.Included from Solo, $69 CAD/month — the way in to the suite

   Assemblies across 20 CSI divisions

   57

   Assemblies across 20 CSI divisions

   Risk rules behind the contingency

   27

   Risk rules behind the contingency

   CAD/month, the way in to the suite

   $69

   CAD/month, the way in to the suite

   Your rates come first

   Every estimating tool that leads with its own cost catalogue eventually gets ignored, because the estimator already knows what their drywall sub charges. So the rate sheet you import is the primary source, always. The bundled Toronto catalogue fills the gaps, validates what you brought, and drives contingency where your data is thin — it never overrules a number you know to be true.

   Feed in completed jobs and it goes further: catalogue defaults get replaced by your crews' measured productivity and cost variance, weighted by how much history it actually has.

   The soft costs nobody models

   On a Toronto residential project the development charge routinely exceeds the entire finishes package, and it is the line most often carried as a guess. The estimator models it properly — permits on the current fee schedule with surcharges, DCs gross and reduced, parkland, the Community Benefits Charge, HST and the new housing rebate.

   Municipal schedules change every year, so every figure is editable in the app, and the jurisdiction it prices against is a data pack rather than a hardcoded assumption — tell it which province, state or country you build in and the rates, tax, holdback and code rules follow. Corrections reach every subscriber the day they ship, with no version to buy.

   A number with a reason attached

   Twenty-seven risk rules produce a recommended contingency, each with a written reason and a mitigation rather than a single opaque percentage. On top sits what-if scenarios, tornado sensitivity, and a Monte Carlo distribution that answers the only question that matters — what is the chance this job loses money.

   The bid strategy layer closes it out: a margin ladder with win probability drawn from the firm's own recorded outcomes, an expected-value optimum, a risk-adjusted floor and a walk-away number.

   • 

   • 

   • 

   • 

   • 

   • 

4. Bid LevelerThe lowest number is not the lowest bid.Included from Pro, $499 CAD/month

   Scope elements across 15 trade packages

   155

   Scope elements across 15 trade packages

   Valuation bases, ranked by strength

   6

   Valuation bases, ranked by strength

   Finding types, from gaps to outliers

   9

   Finding types, from gaps to outliers

   The bid set prices its own gaps

   The interesting move here isn't reading a PDF. It's this: when Sub A excludes crane hoisting and Sub B carries it as a $15,000 alternate, Sub A's gap is worth $15,000. That's a real market price — for this project, this schedule, this site, from a contractor who wants the work. Not a number from a book, and not a percentage.

   So the levelling pass looks for competitor evidence first and only falls back to a local rate table for elements nobody priced. Every adjustment states which basis it used, ranked openly from strongest to weakest: priced by a competing bidder, the bidder's own alternate, your award history, the rate table, a percentage of trade value, or a figure the estimator typed. An estimator who can't say where a number came from can't use it in a negotiation, so the weakest bases are labelled as such rather than blended into a total.

   Silence is the expensive part

   An exclusion is on the record and can be priced. Silence is what reaches site unresolved, months later, with no competitive tension left — so the two are levelled the same way and flagged differently.

   Around that sit the other findings: scope every bidder left out, which doesn't change who is cheapest but does change what the trade costs; allowances wearing a price tag; commercial mismatches like HST in or out, bonding, escalation and validity that expires before award; addenda nobody carried, which makes a bid a price for a different project; price-moving qualifications quoted back with the reason each one matters; and outliers still well below the field after levelling, which in this market is rarely a bargain.

   It produces the day you were about to spend

   The output is the levelling matrix as an Excel workbook — scope down the side, bidders across the top, five sheets, live formulas so an override moves the totals — plus an award recommendation written to be read by a PM who wasn't in the meeting, and one clarification letter per bidder with the reason each open item is being asked.

   It also writes the subcontract scope paragraphs from the levelled position, so what levelling added back can't be quietly re-excluded after signing, and hands its numbers on: JSON to the AI Estimator, CSV to ScopeGuard, a Procore-shaped bid tab for everything else. On the demo package the low bid was $1,142,000; levelled, that bidder isn't the cheapest, and awarding on the bid form alone would have cost $84,986.

   • 

   • 

   • 

   • 

   • 

   • 

5. ScopeGuardThe scope gaps in a buyout, found before award.Included from Crew, $299 CAD/month

   Classes of finding, each with a citation

   8

   Classes of finding, each with a citation

   Trade packages modelled

   20

   Trade packages modelled

   CAD/month, on the Crew plan

   $299

   CAD/month, on the Crew plan

   Reading exclusions, not just inclusions

   The commercially dangerous sentence in a subcontractor proposal is almost never a line item. It is a clause: “excludes winter concreting” on a structure running November to March, or “fire alarm verification by others” on work that cannot be occupied without it.

   The engine reads the master documents and every proposal, then cross-references them — missing scope, required scope explicitly excluded, double coverage between trades, interface grey zones, unusable pricing, and addenda that were never acknowledged.

   Every finding is citable

   A finding nobody can check is worse than no finding, because it costs an estimator an hour to disprove. So each one carries a plain-language explanation, a citation on both sides — the clause in the master documents and the page and line in the sub's letter — a confidence score, and a cost exposure range with its stated basis.

   That structure is what makes the output usable in a buyout meeting rather than just interesting.

   Proven on a real package

   The app ships with a full Toronto buyout — a project manual, three addenda and fourteen real subcontractor proposals — that runs through the same reader, extractor and cross-reference engine a customer's own documents go through. Nothing about the demo is canned.

   On that package it finds, among other things, an excluded fire pump that explains why the sprinkler price looks sharp, an air-barrier tie-in that both the glazing and envelope trades excluded, firestopping that nobody carried at all, and four proposals that never acknowledged Addendum 3.

   • 

   • 

   • 

   • 

   • 

On site, in the pocket

iOS + Android · Q4 2026, included in every plan