Cars24 AI Agent Cuts Task Time 80%, Recovers 12% of Dropout Customers
The case study provides a measurable example of an AI agent handling complex multi-step customer workflows and achieving operational gains beyond simple chatbot automation.
Reporting from 1 source: GIGAZINE.
Cars24, the used car marketplace, deployed an AI agent built on OpenAI's API to handle voice and chat customer interactions. The system now processes over 1 million conversations per month. Support resolution rates improved 50%, and time to complete key service tasks dropped 80%. The agent also re-engaged 12% of customers who had previously dropped out of the sales process. Internally, 600 employees use ChatGPT Enterprise and Codex with 85-90% daily active usage.
Cars24, a used car marketplace in India, faced a familiar problem: transactions that required phone calls, document checks, and follow-ups could take days or weeks. As volume grew, scaling human staff was not sustainable. The company turned to an AI agent built on OpenAI's API that handles both voice and text conversations.
The agent can take a buyer through budget, family size, commute, and vehicle preferences, then recommend cars, schedule test drives, and even handle loan discussions. For sellers, it collects vehicle info, schedules appraisals, and follows up on missed appointments. The results: a 50% increase in support resolution rates, an 80% reduction in task completion time, and the recovery of 12% of dropout customers.
Inside the company, 600 employees use ChatGPT Enterprise and Codex for project management, bug tracking, and finance workflows. Some departments have built their own custom agents connected to Slack, Gmail, and WhatsApp, moving toward a model where teams build their own tools rather than placing requests with engineering.
Synthesized by Yomimono from the 1 cited source below, including Japanese-language reporting where cited, then editorially reviewed before publishing.