How we use AI
Most “AI advisors” for trucking generate their answers directly from a chat model, so their own fine print warns the numbers may be wrong. We built this the opposite way.
The one rule: AI never touches a number.
Every dollar, rate, cost-per-mile, APR, and letter grade is computed by a versioned spreadsheet model — the same kind of workbook a careful CFO would build — recalculated by a real spreadsheet engine. The AI reads your documents, asks for what’s missing, and explains the result in plain English. It cannot compute, adjust, or invent a number or a grade.
What the AI does
- Reads your paperwork. Drop in a rate confirmation and it pulls out the lane, rate, fuel surcharge, and accessorials — then asks you to confirm anything it wasn’t sure about.
- Asks for what’s missing. If a required input isn’t there, it asks — it never quietly assumes a number to fill a gap.
- Explains and advises. Once the math is done, it writes the plain-English “here’s what this means and here’s why I’d be cautious.” That judgment is commentary on the result — it can argue against taking a load, but it can never change the load’s numbers or grade.
What the AI never does
- Compute a cost, rate, APR, cost-per-mile, or grade.
- Change, round, or “adjust” a number the engine produced.
- Make the decision for you. You always decide; we show the math.
You can audit every number
Each figure on your screen carries a source tag — [your cost profile], [you told us], [DOE diesel index], [industry default], or [estimate — range shown] — so you always know where it came from. Behind the scenes every grade traces to the exact spreadsheet cells that produced it, with an assumptions log that flags every default you haven’t confirmed. When we genuinely can’t know a market number, we don’t fake one — we show you the break-even threshold or ask you, and label it.
Same contract in, same grade out
The math is deterministic. Enter the same contract twice and you get the same grade every time — the numbers don’t drift between runs. Only the AI’s wording of the explanation can vary; the figures never do. An independent checker audits every result to confirm the words match the computed cells before you ever see it.
Estimates are labeled estimates
Where a number is an industry benchmark or an estimate, we say so and show a range rather than a false-precision point. Estimates are there to flag outliers against your own numbers, never to replace them.
This is decision-support, not advice
These tools compute and explain; they are not financial, legal, tax, or insurance advice, and they don’t predict insurance premiums. The decision is always yours. Questions about how a specific number was produced? Every result can be traced back to its inputs and the cells that used them.
See also our Privacy Notice, including the firewall promise that your platform data is never used to underwrite or price your insurance.