Agentic AI for Freight Dispatch

Overview

LogiTrack, a mid-market freight brokerage processing 400+ loads per day, needed to automate dispatcher workflows. DevNexus built an agentic AI system that autonomously matches loads to carriers, negotiates rates, and handles check calls via voice AI — increasing dispatcher throughput by 3.2x.

The Challenge

Dispatchers manually matched loads to carriers using spreadsheets and phone calls. Each load required 6-8 phone calls to find the right carrier at the right price. Check calls (status updates from drivers en route) consumed 30% of dispatcher time with zero strategic value.

LogiTrack was growing faster than they could hire. Training a new dispatcher took 6 months, and the labor market for experienced dispatchers was extremely competitive.

Our Approach

1. Load-carrier matching engine: We built a matching engine using historical performance data (on-time rates, damage claims, rate history) and real-time carrier availability from their TMS.
2. Multi-agent system: We designed three specialized agents — one for rate negotiation, one for documentation and compliance, and one for check calls — coordinated by an orchestration layer.
3. Voice AI for check calls: We deployed a voice AI agent (built on Vapi) that calls carriers for status updates, handles natural conversation including delays and exceptions, and logs results directly into their TMS.
4. Human-in-the-loop: Approval gates for loads above $10K, new carriers, and any flagged exceptions. Dispatchers review and approve rather than execute.

The Solution

The system works alongside dispatchers, not instead of them. When a new load comes in, the matching agent recommends the top 3 carriers with predicted rates and on-time probabilities. The negotiation agent handles the outbound communication. Check calls happen automatically every 4 hours via voice AI.

Dispatchers now spend their time on exception handling, relationship building, and high-value strategic decisions — not repetitive phone calls.

Results

• Dispatcher throughput: From 15 loads/day to 48 loads/day per dispatcher (3.2x)
• Carrier rates: AI-negotiated rates saved 22% on average vs. manual negotiation
• Check calls: 91% fully automated, with human escalation for exceptions
• Dispatcher time: 30% of time freed from check calls, redeployed to growth
• Hiring needs: Avoided hiring 20 additional dispatchers ($1.6M annual savings)

Client Testimonial

"We thought we needed to hire 20 more dispatchers. DevNexus gave us an AI team that never sleeps."
— Marcus Rivera, VP Operations, LogiTrack