redBus Builds Multilingual AI Copilots for Bus Operators, Travellers

With AI copilots trending across industries to assist with various tasks, one of India’s leading online bus ticketing platforms has ensured it is not left behind. redBus, founded in 2006 to address bus ticketing inefficiencies, has created an in-house copilot to help their travel partners run businesses smoothly.

Now, their AI copilot for bus operators aims to support their partners by providing them with real-time, data-driven insights that can help streamline operations, improve decision-making, and ultimately enhance the overall customer experience.

During an interaction with AIM, Anoop Menon, chief technology officer at redBus, said, “We feel that this tool can really help us improve our efficiencies externally, internally, and with our customers, and our partners.”

The AI copilot embedded in redPro, a comprehensive platform built specifically for bus operators, is designed to optimise various aspects of their business. Menon emphasised that while the company doesn’t claim to know its operations better than the operators themselves, the AI tool offers “a whole bunch of data, tools, and inferences, to help them manage their business better”.

Actionable insights on route optimisation, inventory management, and revenue growth via historical data and trends analysis are some of the features offered by the copilot.

AI for Regional Markets 

redBus caters to operators across India who speak varied languages by offering multilingual support. The AI tool allows operators to input queries and receive insights in their native language.

“If an operator wants insights in their language, the copilot translates and simplifies the data, enabling them to act accordingly," Menon said. He believes that this functionality can overcome language barriers and ensure that operators can access vital information without technical hindrances.

Notably, big tech players such as Google and Microsoft have developed AI assistants and copilots that can function in multiple languages. Microsoft Copilot is readily available in several Indian languages, including Hindi, Bengali, Tamil and Telugu.

Improving Customer Interaction

Empowering bus operators is one thing, but redBus is also enhancing customer interactions by using generative AI. It has been experimenting with different AI models to determine the most effective solutions for each region.

For example, while OpenAI’s models power chatbot interactions in India, the company is also testing Claude, an AI model developed by Anthropic, in Malaysia. “This gives us the ability to compare outputs and choose the most effective solution for each region,” Menon said.

This AI integration has significantly improved redBus’ customer satisfaction scores, which have risen by 40% since the introduction of the chatbot.

Menon pointed out that the chatbot handles up to 90% of initial interactions while escalating complex cases to human agents when necessary.

At the recent Meta Build Event, which happened in Bengaluru, the company announced partnerships with a number of companies including redBus. 

On the operational side, redBus has also seen improvements in its call centre performance. By incorporating AI into customer service, the company has reduced average call or chat times from an average of seven to nine minutes to just two to three minutes. Internally, redBus uses a code-to-doc system, which generates documentation directly from source code, reducing developers’ workloads and accelerating onboarding for new team members.

redBus is also venturing into voice-enabled booking systems, allowing customers to interact with the platform through voice commands.

AI Challenges Persists

On one occasion, Air Canada’s chatbot provided incorrect information, prompting the airline to compensate the customer. Considering how AI hallucinations still remain a concern, redBus has acknowledged the same. However, to address this, redBus implements validation mechanisms within its AI systems to ensure the quality of responses.

The company implements human review interfaces, combines AI models for cost efficiency, and trains systems to escalate complex cases to human agents to prevent inaccurate responses.

Looking ahead, redBus plans to expand its use of AI across various aspects of its operations, with a particular focus on predictive analytics and dynamic pricing models. “We know that the next big company to challenge us could be a startup. We need to keep innovating and running fast,” Menon concluded.