Perfect human-machine interaction likely in 12-18 months: CoRover CEO

CoRover develops human-centric conversational platforms driven by generative AI (Gen AI) technology, including chatbots, avatars, and voice bots, and has already gained over a billion users

At the NVIDIA AI Summit in Mumbai this year, NVIDIA founder and chief executive officer Jensen Huang and Reliance Industries chairman Mukesh Ambani discussed artificial intelligence (AI) infrastructure investments that will revolutionise industries and position India as a global leader. At the event, many Indian startups also shared their AI-driven  innovations delivering impact in fields such as customer service, sports media, healthcare and robotics.  

 

One of the startups that caught the attention of the developers, academics, and business leaders there was Bengaluru-based startup CoRover.ai. It already has over a billion users of its large language model (LLM)-based conversational AI platform, featuring text, audio, and video-based agents. 

 

CoRover creates human-centric conversational platforms powered by generative AI (Gen AI) technology, including chatbots, avatars, and voice bots for chat platforms. One of its popular products is BharatGPT, a GenAI platform that’s designed specifically for Indian businesses. 

 

“Our vision is to democratise technology by making AI development accessible to everyone. Through a conversational interface, users can create AI applications, such as AI Assistants, by simply describing their requirements,” said Ankush Sabharwal, founder and chief executive officer of CoRover.  

 

“The platform automatically generates text, voice, video, and multilingual content, eliminating the need for technical expertise or coding skills. This aligns with CoRover's broader mission of creating inclusive, intuitive, and human-like AI solutions that empower users regardless of their literacy, language, or technical proficiency,” he added.  

 

For instance, CoRover’s AI platform supports chatbots and customer service solutions for top private and public sector clients. One of them is the Indian Railway Catering and Tourism Corporation (IRCTC), which provides online ticketing and food and water services for India’s sprawling network of railway stations and trains. 

 

It has also formed strategic partnerships with leading global companies, including Google, Bharat Electronics Limited and Mahindra Group. Microsoft has onboarded CoRover as a co-sell partner, ensuring its presence in its marketplace to drive broader market reach and adoption. In fact, the partnership with Google played a pivotal role in the development of BharatGPT. 

 

CoRover differentiates itself further with its multilingual AI capabilities. This allows for seamless communication in regional languages like Marathi, Tamil, Urdu, Punjabi, Assamese, Odia, Bengali, Kannada, Telugu and Malayalam. Its AI agents can perform actions like booking tickets, buying groceries, and processing payments. This transactional capability helps drive customer experience, operational efficiency, and revenue growth for enterprises. 

 

It also powers AskDISHA, the IRCTC’s multimodal chatbot, which handles more than 150,000 user queries daily, and has facilitated over 10 billion interactions for more than 175 million passengers to date. It assists customers with tasks such as booking or cancelling train tickets, changing boarding stations, requesting refunds, and checking the status of their booking in multiple languages, including English, Hindi, Gujarati and even Hinglish, a unique hybrid that of Hindi and English that is widely used in India. 

 

AskDISHA has led to a 70 per cent increase in customer satisfaction for IRCTC. It has also led to a 70 per cent decrease in queries coming through other channels, such as social media, phone calls, and emails. 

 

Sabharwal said the rise of social media and smartphone adoption has resulted in a huge increase in customer inquiries and feedback. This created a scalability bottleneck for traditional call centers, which struggled to efficiently manage the high volume of consumer interactions. Early implementation of AI powered chatbots often fell short in effectively addressing common requests. This was due to limitations in natural language processing (NLP), intent recognition, and contextual understanding. 

 

“This realisation sparked an idea - ‘What if we could create an AI- and NLP-enabled platform that helps companies improve their customer experience’,” said Sabharwal, an alumnus of Birla Institute of Technology & Science (BITS), Pilani. 

 

He founded the firm in 2016 along with Kunal Bhakhri, Manav Gandotra, Rahul Ranjan and Suman Suri. It started with text-based chatbots and gradually expanded to voice and video assistants. The goal was to create assistants that not only provided information but also enabled transactions, offered advisory services, and supported multiple languages and multi-channels 

 

Local data for local needs  

 

The firm created innovative solutions for use cases like booking train tickets, checking PNR (passenger name record) status, transferring money, and comparing prices. Its approach is built on extensive data and insights gathered from collaborations with large enterprises. 

 

What also gives CoRover an edge over larger multinational corporations is access to Indian conversational data in over 14 languages.  

 

“We started this conversational AI journey in 2016, more than a billion users have used our multi-modal AI assistants in various languages. Our AI assistants even work with feature phones,” said Sabharwal. 

 

The platform's technical innovation lies in integrating video, voice, and text into its AI assistants, providing a human-centric experience that mimics natural communication. By prioritising video calls, followed by phone and chat, CoRover enables meaningful interactions, such as face-to-face video meetings. Here users can interact with a virtual representation of a person. 

 

“We demonstrated this during Covid-19 pandemic for doctor-patient interactions,” said Sabharwal. 

 

“I’m optimistic that perfection in human-machine interactions is achievable in 12-18 months, provided we identify the right use cases,” he predicts. “The key is aligning focus with genuine use cases to unlock AI’s full potential.”