HOME CASE STUDIES AI-Powered Call Analytics Platform for Automotive Industry

AI-Powered Call Analytics Platform for Automotive Industry

In this case study, ObjectStyle acted as a development partner of Ander, a provider of intelligent corporate training solutions.

AI-Powered Call Analytics Platform for Automotive Industry
Company Name
Ander
Industry
Corporate Training
Year Founded
2018
Location
USA
Website
ander.ai

Ander serves many B2B clients in the automotive industry. To reimagine how those clients approach call tracking, Ander created Call-iQ, a new-generation call analytics platform that leverages AI for at-scale phone call transcription and analytics.

Challenge

Car dealership agents handle dozens of customer calls every day: inquiries, appointments, follow-up calls, and sales conversations. The quality of those calls directly affects revenue, as well as customer satisfaction and retention. Yet most dealerships have no scalable way to evaluate them. Even with traditional call-tracking, listening to calls takes time, and there are only that many conversations you can go through manually.

As a result, coaching happens reactively, if at all. Managers lack visibility into what’s being said on the phone, agents receive little helpful feedback, and performance issues go undetected until they show up in financial reports.

Solution

Call-iQ automates the entire call evaluation workflow. Every recorded call is transcribed, analyzed against pre-set scoring rubrics, and returned to managers and coaches as a structured report—within minutes and without manual input.

The platform identifies speakers automatically, detects key moments in the conversation, flags concerns that require immediate attention, and generates personalized coaching recommendations for each agent. Managers get a real-time view of team performance across locations and departments.

Since each dealership is different, the platform can be configured to work with scripts for various types of calls and performance metrics to assess those calls against.

What Different Users Do in the System

  • Coaches get automated evaluation across all calls, with no actual listening required.
  • Managers receive team trends, performance leaderboards, and “red flag” alerts.
  • Dealership owners gain visibility into agents’ performance across locations.

The Call Analytics Engine

Here is how the intelligent call analytics platform works under the hood.

Step 1

Recording all incoming and outgoing calls.

Step 2

Using AI for speech-to-text translation and speaker identification.

Step 3

Leveraging an AI reasoning model for in-depth analysis: call categorization and summary, rubric-based performance scoring, “red flag” detection, and coaching action plan generation.

Step 4

Actionable analytics are delivered to the dashboard.

Step 5

Relevant notifications are sent to managers and coaches.

Actionable Data and Reports

Once the data crunching is complete, managers and coaches get the following statistics:
*AI would have analyzed and filtered out empty calls and put voicemail conversations in a separate category.

  • Call summaries
  • Call trends
  • Aggregated scores
  • Top-performing callers
  • Underperforming callers and areas where they are lacking
  • Statistics across dealerships
  • Visual reports with graphs and charts for HR people

Managers can edit AI-produced evaluations by hand, if needed. Since performance criteria and call scripts are in a constant state of flux, new data can be fed to the AI model at any time to start using the updated criteria.

Results

Car dealerships now have an intelligent system, Call-iQ, which they can use for assessing the work of their agents. This saves managers and other stakeholders a lot of time, eliminating the need to actually listen to individual calls. Supervisors can now monitor and assess caller effectiveness at scale, stepping in only when absolutely necessary.

The system is now being used by 20+ dealership clients across the country.

Technology Stack

Backend:

Python

Frontend:

TypeScript / React

Database:

PostgreSQL

Transcription:

Rev AI

AI Analysis:

Anthropic Claude

Storage & Notifications:

AWS (S3, SES), Twilio

Time Span and Resources

Duration:

1.5 years, and counting

Effort:

6250 man-hours