UkurBQAI ESTIMATING · M&E
For M&E subcontractors who price from drawings

Your estimator should be pricing tenders, not counting them.

UkurBQ sets up AI around the way your firm already prices: your BQ layout, your measuring rules, your rates. It drafts the quantities from the tender drawings, shows the sheet behind every line, and leaves the final call to your estimator.

Free test on a tender you've already priced NDA before you send anything Nothing to pay unless you go ahead

Curtis, founder of UkurBQ. I set this up personally for each firm and train your team to use it.

Step 1: see what
you get back

Free BQ test

Book your free BQ test.

30 minutes online. Bring one tender you've already priced and the person who prices it. You get a drafted BQ back within 48 hours to compare with your own.

  1. 1Tell us how you price tenders now
  2. 2We run one of your past tenders
  3. 3Compare it with your own BQ

Step 2: message us
the way that suits you

WhatsApp +60 11-1664 9986
Message on WhatsApp
Email [email protected]
Sound familiar?

You've probably already tried AI on a tender.

Someone uploaded the drawings to ChatGPT. It gave back a list of quantities in seconds. Then your estimator checked every line by hand anyway, because nobody could tell where the numbers came from. It didn't save a single hour.

A general chatbot, used on its own
How many lighting points on this level? (E-11.pdf attached)
Approximately 40–50 lighting points.
Estimated cable: 600 m.
Total M&E cost: RM 180,000.
Which sheet? Which revision? Whose rates? Your estimator can't check it, so they redo it.
  • It doesn't know how you measure

    Every firm counts points, measures runs and adds allowances differently. A general chatbot guesses.

  • It doesn't use your BQ or your rates

    The output still has to be retyped into the consultant's BQ and priced from scratch.

  • It can't show its working

    No sheet number, no revision, no reason when something is unclear. So nobody trusts it, and nobody uses it twice.

The problem was never the AI. It was the setup. That's what we fix.

Where tender time goes

Most of an M&E tender is counting and retyping, whether you win it or not.

Every tender you lose still cost your estimator days of takeoff. These are the hours UkurBQ takes off their desk.

  • Sorting the pack

    Which drawings are current?

    A tender pack lands with dozens of sheets and a closing date. Someone has to sort out revisions before anything is counted.

  • Counting

    Points, fittings, equipment

    Lighting, power and data points tallied floor by floor. FCUs, DBs and fans counted off every layout.

  • Measuring

    Cable, trunking, pipe and duct

    Every run scaled off the drawings by hand, then double-checked because one missed run comes out of your margin.

  • Retyping

    Into the consultant's BQ

    Quantities typed line by line into the BQ or your pricing sheet, then checked again.

  • Addenda

    Doing it all again

    A revised sheet arrives two days before closing. Find what changed, re-count it, update the BQ.

How it works

Read. Count. Your estimator checks. Your rates price it.

The work runs in the same order your team already follows. The difference is who does the counting.

  1. 01

    Tender in

    Upload the drawing set and the consultant's BQ.

  2. 02

    Read and sort

    Sheets, revisions and schedules are picked out.

  3. 03

    Count and measure

    Every quantity drafted with its sheet and revision.

  4. 04 · YOUR ESTIMATOR

    Check and approve

    Marked lines reviewed. Nothing moves on without sign-off.

  5. 05

    Price

    Your rates applied in your own BQ layout.

  6. 06

    Addenda

    Revised sheets compared, changed lines listed.

It runs inside your own Claude or ChatGPT business account, set up around your templates. You don't learn new software.

Proof before you rely on it

We prove it on tenders you've already priced.

You don't have to trust a demo. Before your team uses it on a live tender, it has to match work you've already done.

BEFORE YOU PAY

The free test

One past tender, run once. You compare our draft with your BQ and decide whether it's worth setting up.

JOB 1 · TOGETHER

Run it line by line

We go through a priced tender with your estimator and adjust the setup wherever it differs from how you measure.

JOBS 2 & 3 · BLIND

Lock it, then test it

Two more past tenders, run without changes. Only when they match your own BQ do you use it on a live tender.

Keep quoting the way you do now until the proof runs match. Nothing changes for your team until you're satisfied.

Sample output

Every quantity carries its sheet and revision.

A drafted BQ for one floor of a mixed electrical and ACMV package. Where the drawings disagree or can't be scaled, the line is marked for your estimator, with the reason.

RefDescriptionQtyUnitSheetStatus
E1/03Lighting point wired in 1.5mm² PVC cable in PVC conduit46ptsE-11 Rev B● Counted
E1/07Emergency light, self-contained, 3-hour9nosE-11 Rev B● Counted
E2/0213A switched socket outlet point wired in 2.5mm² PVC cable32ptsE-21 Rev B● Counted
E4/01Submain 4C 25mm² XLPE/SWA/PVC to MSB-2
Length scaled from a sheet marked "not to scale". Confirm against site route.
64mE-31 Rev A▲ Check
M2/04Ceiling cassette fan coil unit, 2.0 HP, as schedule8nosM-03 Rev A● Counted
M3/02Insulated refrigerant piping to condensing unit CU-3
Outdoor unit position differs between M-03 and roof plan A-09.
22mM-03 Rev A▲ Check
M5/01Kitchen exhaust fan, as schedule2nosM-07 Rev A● Counted

7 lines · 5 counted · 2 to check · Sample data from a fictional project. No prices shown.

Compared

Why a setup beats a chatbot and beats doing it by hand.

UkurBQ setupChatGPT on its ownBy hand
Shows the sheet behind each quantityYes, every lineNoOnly in the estimator's head
Uses your BQ layout and measuring rulesYes, set up for youNo, re-explain every timeYes
Flags unclear or clashing drawingsYes, with the reasonNo, it guessesIf someone notices
Handles addendaChanged lines listedStart againRe-count by hand
Estimator signs off before pricingBuilt inUp to youYes
Time on a typical takeoffHours, mostly checkingMinutes, then a full re-checkDays
We train your team

A tool nobody knows how to use is worth nothing.

Training is part of the setup, not an extra. Your team leaves able to run a tender on their own, and a new hire can learn it without us.

  • Hands-on session

    We run a real tender with your estimator: loading the pack, checking marked lines, handling an addendum.

  • Written guide

    A short step-by-step guide written for your setup, your BQ layout and your rules.

  • Recorded walkthrough

    A screen recording of the whole process, so knowledge doesn't leave when someone does.

  • Support afterwards

    Questions answered on WhatsApp, and adjustments when your rates or templates change.

What you get

Everything set up, tested and handed over.

  • BQ drafting set up for your trade: electrical, ACMV or solar
  • Your BQ layout, measuring rules and rate sheet built in
  • Sheet and revision on every quantity
  • Addenda comparison and change list
  • Proof runs on three of your past tenders
  • Tender tracker for closing dates and supplier quotes
  • Hands-on training, written guide and recording
  • Runs in your own AI account; you own the setup
Price One fixed setup fee

Agreed on the call, after we've seen how you price, and before you commit to anything. Your Claude or ChatGPT business plan is paid separately, directly to them.

Book a free BQ test
Founding firms

We're taking on our first M&E subcontractors now. Founding firms get hands-on attention from the founder and pilot pricing in return for honest feedback.

FAQ

Straight answers.

We tried ChatGPT and it was way off. Why is this different?

A chatbot on its own doesn't know how you measure, what your BQ looks like or how you price, and it can't show where a number came from. UkurBQ is set up around your rules and shows the sheet and revision behind every quantity, so your estimator can check it instead of redoing it.

Will it replace my estimator?

No. It takes the counting and retyping off their desk. They still check the marked lines, set the rates and decide the price. Most firms use the time to price more tenders.

How do I know the quantities are right?

It's proven on three tenders you've already priced before you use it live. On every job after that, each quantity shows its source, and anything it can't confirm is marked for your estimator rather than guessed.

Are our drawings and rates safe?

We sign an NDA before you send anything. Your setup runs in your own Claude or ChatGPT business account, and your files and rates are never used for another firm. Test files are deleted on request.

Does my team have to learn new software?

No. It runs in an AI account your team can open in a browser, and the output follows your existing BQ layout. We train your team in a hands-on session and leave a written guide and a recording.

Do we have to set it up ourselves?

No. We map how you price, build the setup, test it on your past tenders and hand it over. Your part is answering questions about how you work and sending past tenders.

Can it work with the consultant's BQ format?

Yes. Send the blank BQ from the tender documents and the draft follows its item numbering and units.

Which trades do you cover?

Electrical, ACMV/mechanical and solar. Working in another trade? Ask anyway.

What does it cost?

One fixed setup fee, agreed on the call before you commit, plus your own AI subscription paid directly to Anthropic or OpenAI. The first test is free.

Where are you based?

In Malaysia. We work with subcontractors across the country by WhatsApp, email and video call.

Next step

Send one tender you've already priced.

See our draft next to your own BQ within 48 hours. Free, and no commitment.