Skip to content

Your guide

Where AI Actually Fits in Mechanical Contractor Purchasing

Stop paying your best people to chase quotes. See what AI can take off their plate, and what should always stay in your hands.

Thanks for grabbing this.

Quotes arrive as PDFs, Excel files, emails, and screenshots. Someone retypes them into a sheet. Then the detail that costs you is the one nobody catches: a lead time that quietly grew, payment terms that changed, a note on page 3 saying the price expires. This guide shows where that time and money go, and where AI actually fits.

I'm Hassan, founder of Creativecodes. I build AI-assisted supplier RFQ workflows for commercial HVAC and mechanical contractors.

What's inside

  1. A 2-minute way to price what quote chasing costs you
  2. The 4 steps of an RFQ: where each one leaks and what you still decide
  3. The one rule: prepare, flag, approve
  4. When you don't need this
  5. How to review one past bid.

What is quote chasing costing you? (2-minute math)

Quotes per month×hours per quote×your team's hourly cost

That's what you pay before anyone makes a single decision.

Now add two more numbers:

  • What did the last easy-to-miss detail cost you? A lead time that slipped, payment terms that changed, a scope line left out.
  • How many bids has a late quote put at risk?

What you could get back

Time

Your team stops chasing replies and retyping quotes.

Money

Changes in lead time, payment terms, and scope show up before the PO, not after.

Speed

One side-by-side comparison, ready for review, instead of five attachments and an email thread.

Calm

Everyone can see who replied, who declined, and who needs a nudge.

Want to see what that looks like for your team? Book a 30-minute call → (opens in a new tab)

How it works

  1. Step 1: Find suppliers and prepare the RFQ

    The leak
    Your team hunts for contacts and rewrites nearly the same email for every job. One wrong drawing revision attached to the RFQ goes to every supplier at once.
    What the AI does
    Pulls likely vendors from your supplier list and drafts the RFQ from your project details.
    Result
    Your team starts from a finished draft, not a blank email.
    You decide
    Your team approves the suppliers, scope, quantities, and drawings before anything goes out. If a spec is missing, the system asks. It never fills the gap.
  2. Step 2: Follow up without the nagging

    The leak
    Someone has to remember every open request and ask, "Did anyone hear back?"
    What the AI does
    Tracks replies and follows up by email, SMS, or phone, using rules you set.
    Result
    Fewer late quotes, and nobody keeps a mental list.
    You decide
    A supplier who says "we can't hit that date" or offers a substitution goes to a person, not another reminder.
  3. Step 3: Read every quote, whatever the format

    The leak
    PDFs, Excel files, emails, and screenshots all get retyped by hand. Units and descriptions never match. Lead time, payment terms, and how long the price is good for sit in a different place in every quote.
    What the AI does
    Pulls out prices, quantities, lead times, payment terms, quote validity, freight, inclusions, and exclusions.
    Result
    No more copy-paste into comparison sheets.
    You decide
    Missing information is never guessed.
    What the supplier saidWhat you see
    Lead time is statedShown as stated
    Payment terms differ from your RFQFlagged: differs from RFQ
    Quote validity isn't mentionedNot stated, confirm
    A price is unreadableUnable to verify, review source
  4. Step 4: Compare side by side

    Example (made-up quotes)

    Supplier ASupplier B
    Lower totalHigher total
    Lead time not statedLead time: 3 weeks
    Payment terms: 50% upfrontPayment terms: net 30
    Alternative item offeredSpecified item quoted
    The leak
    Supplier A looks cheaper, until someone finds out the lead time isn't stated, the payment terms are tougher, and the item isn't what you specified.
    What the AI does
    Lines up the differences and flags them.
    Result
    Your team spends time on the decision, not on finding the gaps.
    You decide
    Is the alternative acceptable? Do the payment terms work? Does the lead time fit the schedule? The AI never picks the winner.

The one rule: prepare, flag, approve

Prepare

AI builds the request and organizes the responses.

Flag

AI points out anything missing, unclear, or conflicting.

Approve

Your people approve what goes out and decide what gets bought.

Supplier selection, negotiation, technical acceptance, and purchase approval always stay with your team.

When you don't need this

If your current software already does this well, or you only handle a few quotes a month, don't add another tool. I'd rather tell you that on a call than sell you something you won't use.

Not sure which side you're on? Let's check on a 30-minute call → (opens in a new tab)

Start with one past bid.

Pick a job your team has already priced. Look at the original request, the supplier replies, and the comparison your estimator put together.

  1. Follow the request.

    What needed pricing? Which suppliers did your team contact, and what information did they need?

  2. Review what came back.

    Where did someone chase a reply, retype a price, or check freight, lead times, exclusions, or a different model?

  3. Choose one step to improve.

    Start with a recurring task that takes time. Agree on what the workflow should prepare, what it should flag, and who reviews it.

You don’t need a new process mapped out before we talk. A walkthrough of one past bid is enough to start. Keep confidential bid documents off this website.

Want to see what this would look like for your team?

Book a free 30-minute call. We'll look at how your team handles quotes today, and you'll leave with:

  1. A clear picture of what could come off their plate
  2. A list of what should stay with your people
  3. An honest answer on whether a small pilot is worth trying

No slides. A rough idea of your quote volume is enough.