Zoomcar taps Google Cloud to integrate GenAI and ML for better CX

Zoomcar Holdings, one of India’s largest self-drive car rental platforms, on Wednesday, collaborated with Google Cloud to integrate Generative AI (GenAI) and Machine Learning into its platform in a bid to improve guest bookings, onboarding and the safety and security of clients.

The announcement comes at a time when the company is reeling from the crisis of a cyber attack.  Notably, in June, the company reported a cybersecurity incident affecting 8.4 million users, whose names, phone numbers, vehicle registration details, personal addresses, and email addresses were reportedly exposed. The company, however, said it has isolated affected systems, enhanced network monitoring, and is initiating security audits of its cloud infrastructure. Third-party cybersecurity experts have been assisting the investigation.

Nonetheless, the Bengaluru-headquartered company said that it has made significant progress in automation and AI-led customer experiences, many of which are now being further scaled through the partnership with Google Cloud.

Over the past three months, Zoomcar has used Google Cloud’s Vertex AI and Gemini models to deploy intelligent AI agents across the user journey. These solutions are delivering improvements in user experience, operational efficiency, and trust on the platform.

One of the innovations is Host Assist—a Large Language Model (LLM)-powered AI assistant designed to simplify and accelerate the host onboarding journey. Hosting a car on Zoomcar comes with its nuances, and Host Assist clarifies the process by delivering instant, policy-aware guidance 24×7 to prospective hosts.

Built on a Retrieval-Augmented Generation (RAG) framework and leveraging the Gemini Flash 2.0 model, Host Assist has helped prospective hosts complete their onboarding journey faster—driving a 30% reduction in average onboarding time. By providing real-time query resolution, it eliminates the need for hosts to wait for live interactions with Zoomcar’s Host Success team.

To support hosts beyond onboarding, Zoomcar is building a Host Intelligence Engine—an AI-powered platform that delivers personalised feedback and actionable insights to help maximise host earnings and lifetime value. By analysing host behaviour, actions, and benchmarking against top performers, the engine offers tailored recommendations on key business levers. Zoomcar has begun with Pricing Insights—providing data-backed suggestions to help hosts price their cars more effectively and improve earnings.

As part of its AI-led efforts to enhance marketplace quality, the company has also focused on reducing host cancellations. By leveraging Google Cloud’s Gemini models to analyse host-guest chats and voice calls, Zoomcar can identify responsibility in cancellation disputes.
This enables timely feedback and corrective actions with hosts, leading to improved behaviour. This is expected to reduce host-driven cancellations, directly contributing to a better guest experience and repeat rates.

To enhance guest quality, Zoomcar is also using a real-time AI-driven fraud detection engine built on its Customer Data Platform (CDP). By analysing encrypted identity attributes, KYC documents, and selfies, the system proactively flags suspicious activity. This equips the operations team with timely alerts and is expected to improve fraud detection accuracy in future, further enhancing platform integrity and host confidence.

Vishal Ramrakhyani, Head of Product & Engineering at Zoomcar, said their focus is on applying GenAI to create real-world impact for customers and internal teams. He noted that with Google Cloud, they've built AI agents and ML models deeply embedded in their workflows, driving commercial results and building customer trust and loyalty.

With current annual revenues of $9.11 million, Zoomcar reports that its platform includes over 40,000 vehicles and serves more than 10 million users across 90 cities.