Cars24 recovers 12% of lost leads with OpenAI voice agents

Car buying in India is famously high-touch — it bleeds across calls, document checks, and weeks of follow-ups. That friction used to cost Cars24 a chunk of its seller pipeline. Now the company says AI agents built on OpenAI’s APIs are recovering 12% of leads that previously went cold after 10 days, part of a broader deployment that now handles over a million conversation minutes every month.

Those aren’t just chatbot pings. When a buyer calls, an agent probes for budget, family size, and commute needs, then recommends inventory and books test drives. It confirms appointments beforehand and circles back afterward to gauge intent. Sellers get a parallel thread: the agent collects vehicle details, schedules inspections, and re-engages people who bailed — qualifying whether they’re back in the market at a price Cars24 can hit. The company reports a 50% jump in support resolution rates and an 80% reduction in turnaround time across key service workflows.

What’s unusual is how deeply the tooling has bled beyond the customer-facing layer. Cars24 rolled out ChatGPT Enterprise and Codex across engineering, finance, legal, and marketing. Product managers now use Codex to groom Linear tickets. Engineering tags it into bug reports to pick up defined tasks. Finance teams pull numbers from systems of record and prep investor reports without chasing business heads for inputs. “I thought Codex would make our engineers faster,” one leader noted. “What surprised me was how quickly it spread beyond engineering.”

That spread forced the company to reorient project management around Linear in weeks, creating a cleaner path for AI to participate in day-to-day work. The bet isn’t just on automation — it’s on changing how the whole org thinks about getting work done. Whether that cultural shift sticks when the novelty fades is the real test. For now, Cars24 is treating AI less as a tool and more as an operating layer, and the numbers suggest it’s paying off.