PaveOps AI: Paving Estimates Where the Math Is Math

In paving, estimating happens after the crews go home.

RFQs show up by email, by PDF, and by phone. Somebody has to pull out the scope, check square yards and tons, run the numbers, write the proposal, and then chase it. A wrong depth or a missed area costs real money. And there's rarely time for a second look.

That's the problem PaveOps AI is meant to help with.

What PaveOps AI is

PaveOps AI is an estimating tool I'm building at Dreki Solutions for paving contractors. It's in development. It is not finished, and it isn't for sale yet. I'm building it with a small group of founding pilot contractors in the Midwest.

Estimating comes first. The goal is to help move an RFQ from intake to a checked estimate and proposal, with a person approving every estimate before it goes out.

Why "the math is math"

Here's the part I care about most. The quantity and pricing math should not be AI guesswork.

AI is useful for reading an RFQ and pulling out the scope. It is not what should be computing your tons. That should be standard formulas and your rates. Same inputs, same number, every time. If your estimator wants to know where a number came from, the answer should be a formula, not "the AI said so."

How it's supposed to work

This is what we're building toward. Pilot feedback sets the order.

1. RFQ intake. Requests land in one place instead of scattered across inboxes.
2. Scope extraction. It pulls areas, depths, mix, and site conditions out of the RFQ for you to review.
3. Quantity check. Square yards, tons, and other quantities get checked against the plans before anything is priced.
4. Estimate. Your rates, standard formulas. You approve it.
5. Estimate audit. A second pass flags missing items and numbers that look off before the estimate leaves the office.
6. Proposal and follow-up. The approved estimate becomes a proposal, and follow-ups don't slip.

Further out, we may look at production planning, trucking, and pre-job packets. That's later, and only if pilots say it matters.

Why a second look matters

Most estimating mistakes aren't hard math. They're a missed area, an old addendum, or a depth that got copied wrong. The estimate audit step is meant to catch those. It compares what the RFQ asked for with what the estimate covers, and flags anything that doesn't line up. Your estimator decides what to do about each flag.

Why build it with pilots

A small group of paving contractors gets in early. In return, they give a short call every couple of weeks, straight feedback, and a few past RFQs and estimates, scrubbed if they like, to test against. They see each piece before it's released and help decide what's next.

I'm not going to tell you what it will save. It isn't built, and nobody has run a real RFQ through it yet. When there's something real to show, I'll show it.

Who's building it

Dreki Solutions is a veteran-owned software company in Michigan. You'd work directly with me, Brett Moser, the founder.

Next step

Want your estimating process built in from the start? Call (517) 215-7573 or email brett@dreki-solutions.com with "PaveOps AI" in the subject. Tell me the work you bid and how you estimate today. I'll tell you where the build stands and how the founding pilot works.