Case Study · Freight & Logistics

United Freight & Transport: Dispatch That Runs Itself

UFT moves freight across Alaska — a state where a single missed handoff can mean a load sitting on a dock for three days. They came to Attu because their dispatchers were spending more time reconciling paperwork than dispatching loads.

The problem

UFT had grown the way most regional freight outfits grow: faster than the back office could keep up with. Manifests came in by email, by PDF, and by phone. Dispatchers were retyping the same data into three systems, chasing missing paperwork from drivers, and reconciling weights and stops by hand at the end of every shift.

Two things were happening as a result. Dispatchers were burning out — the job was supposed to be operational decisions, not data entry. And the back office couldn't tell on a given Tuesday afternoon how the week was actually trending without manually pulling reports.

What we built

We started in the dispatch room. Two weeks of sitting with the team, mapping how a manifest actually moves from inbound email to driver app to billable invoice.

The system we built ingests inbound manifests in whatever form they arrive — email body, PDF attachment, scanned document. It extracts the fields, classifies the load, checks for completeness, and pushes the dispatchable record into UFT's existing system. Anomalies route to a human; everything else routes itself.

On the back end, the same system produces the daily and weekly rollups that ops leadership was building by hand on Friday afternoons.

The outcome

"Attu came to us, sat with our dispatchers, and built something that actually fit how we work. Within weeks our team was using it on their own — no hand-holding needed." — Samantha, United Freight & Transport

Why it worked

Two reasons, mostly. First, we built against the work as it actually happened, not the work as it appeared in the org chart. The dispatchers had workarounds for every edge case in the business; the system had to honor those workarounds, not pretend they didn't exist.

Second, we stayed through adoption. The build was the cheap part. Making sure a freight team in production trusts a new system, and uses it instead of working around it, was where most of the work lived.

This is what we mean by workflow automation. For the broader story, see AI automation for Alaska businesses.

Ready to put AI to work?

Tell us what's eating your team's week. We read every inquiry personally.

Get in Touch →