The only AI in trucking that gets better at your book every week.

RadarIQ has read the file on every active US carrier and private fleet running ten or more trucks — safety record, insurance history, fleet moves, the software they run. Ask it for the accounts you want the way you would ask a colleague and it builds the list. Then it does the thing a database cannot: it learns what you sell from the work you actually do, and re-ranks the whole universe around it. A filter does the same thing on day 200 that it did on day one. This does not.

Ask in a sentence. Get a list you can work today.

There is no query builder and no saved-search syntax. You describe the account the way you would describe it out loud, and RadarIQ decides which of the fifty-odd signals on file actually answer the question.

Real exchange from the platform, unretouched. Vendor names removed.

Read what it did there. Nobody told it that a missing vendor record is evidence of manual recruiting — it reasoned from an absence in the data, which is the kind of inference a filter cannot express at all. Then it ranked on the signal that makes the pain acute rather than filtering on it, and offered to go verify before the rep spends a call finding out.

It asks you questions back

This is the part that has no equivalent anywhere else in trucking. RadarIQ is not parsing your sentence into a WHERE clause. It is holding a working conversation about how you sell, and it pushes on the parts it does not understand yet.

Real exchange. Vendor names removed; nothing else changed.

No tool in this market asks a rep whether a talking point is the opener or the closer. That question is worth more than the list, because the answer changes how RadarIQ writes every brief and every draft it hands that rep from then on.

What it actually returned

The Texas pull came back as twenty-five carriers staged for review. You can judge the family-owned read yourself from the names it surfaced:

Real query, real leads, unretouched. Strain and Sons. Nelson Bros. Pfeiffer and Son.

Two learning loops, and neither one has a setup screen

Most tools that claim to personalise ask you to fill in a form first, then never change again. RadarIQ learns twice over, continuously, from things the rep was going to do anyway.

Here is the profile RadarIQ had written about the recruiter above, in its own words, after a handful of conversations — alongside the counter of what its behavioural loop had absorbed separately:

Written by the platform, not by the rep. Behavioural loop alongside it: 11 good fits, 0 skips, signal learning active.

“Growth signals are the opener; tech stack mismatch is the closer” is a sales strategy, and nobody typed it in. RadarIQ inferred it, wrote it down, and now ranks against it. A new hire inheriting that territory inherits the reasoning with it.

Eight ways of selling, one carrier universe

RadarIQ trains a separate profile per persona — recruiting, safety and compliance, TMS and fleet tech, insurance, brokerage, maintenance, factoring, toll management — and a rep can hold more than one. Two people in the same company looking at the same carrier get different reasons in a different order, because they are different reasons.

An insurance producer asking about a dated cancellation gets renewal position, incumbent carrier and the markets that would write the risk. A recruiter asking about the same fleet gets hiring signals, out-of-service tells and which ATS is on file. Same record, same day, two answers.

What it will not do

An AI that sells into trucking has to be trusted on dates and numbers, so RadarIQ is built to decline rather than guess. It will not print a renewal date the federal record does not support — a BMC-91X filing never expires, and where the record genuinely does not say, it says so instead of adding twelve months to an old filing. It will not turn a crash count into a risk verdict without the fleet size underneath it. And it does not push anything into your pipeline; every list it builds is staged for your review, and only what you send reaches RadarHQ or your CRM.

The rest of the stack runs off the same profile. RadarResearch writes a cited brief on any account it staged. RadarOutreach drafts the email, the LinkedIn note and the voicemail with your angle already in them. RadarRoadshow groups the ones worth a visit into a day by territory. RadarNetwork checks whether anyone on your team can already get you in the door.