Havells has launched an Agentic AI Voice feature in its Havells One app that allows users to control connected appliances through natural speech in English and seven Indian languages.
The feature is built by Havells Centre for Research & Innovation (CRI), with ElevenLabs providing the streaming Speech-to-Text and Text-to-Speech models for the voice interface.
The feature is live for all 2.67 million Havells One mobile app users and supports Hindi, Tamil, Telugu, Bengali, Kannada, Gujarati and Malayalam, in addition to English.
Users can use voice commands to control appliances, rooms, scenes and coordinated actions across connected devices. The feature currently works across Havells fans, air purifiers, water purifiers and water heaters, as well as Lloyd air conditioners and washing machines, Crabtree smart switches and upcoming smart home products.
Unlike Havells’ earlier voice experience, which relied on predefined commands and supported English and Hindi on selected appliances, the new system is designed to understand user intent. It can also handle code-mixed speech, where users switch between languages while speaking.
For instance, a request to “make the room comfortable” can trigger coordinated actions across an air conditioner, air purifier, fan and lights. Requests involving multiple devices can be executed across different connectivity protocols.
The agent also retains context across conversations and can ask users to clarify ambiguous requests. It can explain device error codes in spoken language and is being extended beyond device control to areas such as product FAQs, troubleshooting and scheduling.
“This Agentic AI Voice capability is a breakthrough for our customers, allowing them to control and interact with their devices naturally, without having to remember commands or learn rules,” said Dipesh Shah, Executive President and CTO, Havells.
“Its potential is limitless and will continue to expand as new capabilities are added across our devices and services. As a daily-use AI experience, it showcases the transformative potential of AI agents, far beyond the traditional voice interfaces of the past,” he added.
Havells said its CRI team evaluated voice AI models based on recognition accuracy, multilingual support, latency and voice synthesis. The company moved the solution into production in approximately six weeks, with support from ElevenLabs’ Forward Deployed Engineering team for voice quality and regional language pronunciation tuning.
“People have always had to learn the interface of communicating: the commands, the syntax, the order they come in. That is hardest in India, where a household does not speak one language and often moves between three in a single sentence,” said Karthik Rajaram, General Manager and Country Head, India, ElevenLabs.
“Powered by ElevenLabs, Havells now runs seven Indian languages alongside English across an appliance range, so the person no longer has to meet the technology halfway. They say what they want to the agent, in the language they were already thinking in, and the devices respond,” he added.
Havells plans to expand the feature to more Indian languages and bring additional appliances onto the voice platform. The company also plans to add greater personalisation for individual household members and extend voice interactions to product support, troubleshooting and scheduling.
- Europe's Biggest Tech Startup Cuts AI Coding Costs by 90% Using Chinese Model
- E2E Networks Launches Blaze To Offer Pay-Per-Token AI Inference Services
- Apple Posts Record Revenue in India for Q4 FY26
- Dispatch Network Raises ₹5 Cr Pre-Seed Led by Antler India to Build AI-Powered Delivery Network
- Coforge Launches BridgeCX With Pega, Mastercard for Customer Personalisation
- HCLTech Launches AI Agents to Support Mainframe Operations

