← Cases
United StatesProfessional services2025 – ongoing

Cox Automotive

FleetMate assistant for fleet repair estimates

Cox Automotive cut fleet repair estimate preparation from 8 to 48 hours to 30 minutes with AI agents

WorkingVendor case studyVendor ↗

Problem

Complex fleet repair estimates took 8 to 48 hours to prepare. Cox also lacked a standard environment for running agents with appropriate security and monitoring.

Approach

Cox built FleetMate as an internal fleet services solution and deployed it on Amazon Bedrock AgentCore. Its estimator assistant automated estimate creation, using a shared environment for agent execution and access controls.

Results

8 to 48 hours to 30 minutestime to prepare complex fleet repair estimates, per AWSVendor case study[1]

Takeaway

Cox reported a production result for estimate preparation while its driver-call agent remained in a pilot.

Categories

Business function
Production and quality
Innovation type
Process innovation
Evidence
Vendor case study

Sources

Source summaries

[1]Amazon Web Services5 Dec 2025

AWS described Cox Automotive's production deployment of FleetMate on AgentCore. It reported that complex fleet repair estimates took 30 minutes instead of 8 to 48 hours. The page carries no date; the Internet Archive first saved it on December 5, 2025.

Full details are in the original sources.

Added 4 Nov 2026 · Updated 4 Nov 2026
Open case page ↗