Cars24 turned what in India is usually a long, fragmented process into a faster, more consistent experience thanks to AI. Can you imagine coordinating financing, visits, documents and follow-up without growing a human team at the same rate? That's exactly what they did: put voice and chat agents to handle conversations and processes that used to take days or weeks.
How Cars24 uses AI in the buying and selling journey
Cars24 operates one of the largest AI-native automotive ecosystems in the world, with presence in India, the United Arab Emirates and Australia. In a market where buying and selling rarely happens in a single session, the company started by solving the mid and lower funnel: the places where conversation drives conversion.
Using OpenAI APIs, Cars24 deployed voice and chat agents that guide the customer from the first call through post-sale service. When you call as a buyer, an agent asks about budget, family size, commute time and type of car, suggests options from the catalog, schedules test drives and helps with financing. Before the visit it confirms details, captures preference changes and prepares financing data. After the test drive it re-evaluates interest, offers alternatives and keeps the follow-up going.
For sellers, agents collect vehicle details, schedule inspections, send reminders and rebook missed appointments. And for leads that used to go cold after 10 days, the AI re-engages, reassesses renewed intent and brings customers back into the funnel when the price makes sense.
Result in numbers: more than 1M+ minutes of conversation per month handled by AI, 50% increase in support resolution rates, 80% reduction in key response times and 12% of seller leads recovered.
AI as an internal layer: ChatGPT Enterprise and Codex
The move didn't stop at the customer-facing side. Internally, Cars24 rolled out ChatGPT Enterprise and Codex across about 600 employees in the central organization, reaching 85% to 90% daily usage. What does that mean day-to-day?
- Product teams use
Codexto create and refine tickets in Linear. - Engineering tags tasks so
Codexcan take defined steps: from implementation to bug resolution. Codexsummarizes activity in GitHub and publishes updates, reducing the need for so many standups.
The surprise was that Codex spread beyond code: product, finance, legal and marketing began building their own flows. In finance, for example, Codex extracts numbers from ledgers, runs analyses and prepares investor reports without manually chasing inputs.
Another automated flow reviews purchase requests above a defined threshold, detects anomalies, flags risks and can auto-approve when no issues are found.
Change how you work, not just speed up tasks
What's notable is the cultural shift: teams that used to wait on central support now build their own tools. Some created agents like a 'chief of staff' that connect Slack, Gmail and WhatsApp to handle communications, scheduling and follow-ups.
Cars24 shows that AI scales when it's tied to concrete business flows — not as loose experiments — and when it's integrated into both the customer experience and internal operations. The result: broader lead coverage, more consistent processes and the ability to grow without multiplying teams at the same rate.
So what's next? More personalization, better recommendations and an experience where AI keeps service coherent without losing the human touch when it's needed.