Documentation → Best Practices

Best Practices

Playbooks from real carrier operations — how teams use AI dispatch, feedback loops, and alerts to keep trucks moving and drivers happy.

AI Optimization

Training the AI with feedback

Your thumbs up/down on recommendations isn't just cosmetic — it directly improves matching accuracy for your fleet.

Key practices

  • Rate every recommendation. Even when you override, tell the AI why. "Wrong equipment" trains differently than "driver unavailable."
  • Update driver preferences regularly. When a driver says "no more California runs," add it immediately. The AI only knows what you tell it.
  • Check accuracy weekly. The AI dashboard shows your acceptance rate trending over time. Most carriers see 10-15% improvement in the first month.

"We went from overriding 40% of recommendations to under 15% in six weeks just by being consistent with feedback."

— Operations Manager, 35-truck fleet
Driver Retention

Setting up driver preferences right

Drivers who get loads that match their preferences stay longer. Here's how carriers configure profiles for best results.

What to capture

  • Home base and radius. Where do they want to end up? How far will they go? This single field eliminates 30% of bad matches.
  • Preferred and excluded regions. Some drivers love the Northeast corridor. Others won't touch I-95. Capture both.
  • Load type preferences. Drop-and-hook vs live load, hazmat certified, oversize experience — the details matter.
  • Schedule constraints. Days off, appointments, recurring home time. The AI won't offer loads that conflict.
Efficiency

Getting the most from OCR parsing

The AI reads rate cons fast, but small habits make it even faster and more accurate.

Pro tips

  • Use digital PDFs when possible. Scanned documents work, but native PDFs from broker portals parse with higher confidence.
  • Always verify multi-stop sequences. The AI usually gets it right, but stop order is worth a quick check before dispatch.
  • Check appointment times on handwritten docs. Scanned handwriting is the one area where OCR needs extra verification.
  • Batch similar rate cons. Processing 5 loads from the same broker? The AI recognizes the format and speeds up.
Operations

Using alerts to stay ahead of problems

Smart alert configuration means fewer fires. Here's how carriers set up notifications that actually help.

Alert priorities

  • Unassigned loads approaching pickup. Get notified 24 hours before — not the morning of.
  • Missing documents on delivered loads. Chase PODs before they're 48 hours late, not a week.
  • Drivers with incomplete profiles. Catch new drivers before they get offered loads the AI doesn't have enough data to match.

Have a playbook to share?

We're always learning from carriers. If you've developed workflows that work well, we'd love to hear about them.

Share Your Story