About Meesho

Meesho is a leading Indian social commerce platform that enables over 18 million entrepreneurs, mostly small entrepreneurs and stay-at-home mothers, to resell on WhatsApp, Facebook, and Instagram. It was established in 2015 by Vidit Aatrey and Sanjeev Barnwal. The company has experienced unprecedented growth, with over 170 million users on a monthly basis in 2024. Growth is one of the reasons resulting from the fact that Meesho is extremely well established in Tier II and small cities, and its zero-commission facility provides convenience and affordability for sellers to start their ventures.

To keep up with growing user scale and offer frictionless support at scale, Meesho introduced a multilingual voice bot in November 2024, the first from Indian e-commerce. The voice service is revolutionizing the resolution of support queries at its very core, minimizing wait time, decreasing cost, and providing a smoother experience for the users in the country.

Why Meesho Required an AI Voice Bot

With the growth of Meesho, customer engagement increased in terms of intensity and numbers. Certain challenges had to be automated:

• Huge Amounts of Support Requests: With hundreds of thousands of requests per day, human reps were bogged down, with delays and lost business opportunities created. Now, the bot handles 60,000+ customer calls per day, with 24×7 support.
• Language Barriers: 70% of Meesho’s clients communicate in local languages, and that has to be facilitated with multilingual support at scale. The bot is currently supporting English and Hindi, with six other Indian languages supported.

• Query Resolution: Human agent-only resolution costs too much. Shifting to automation gave a 95% resolution rate and minimized human intervention to a large extent.

Technology Used

Meesho uses a GenAI-voice bot, which is based on cutting-edge technologies like text-to-speech (TTS), natural language processing (NLP), automatic speech recognition (ASR), and large language models (LLMs) to offer real-time, accurate and context-specific customer care. The application allows the bot to make subtle interactions like detecting colloquial approvals like “yes,” “ji,” or “okay” and interruptions so that the conversation flows without any interruptions. In the near future, Meesho is going to incorporate emotion recognition features so that the bot can detect the emotion and tone of a user and respond more empathetically. This shift is going to make AI-driven support more human-like so that experiences are felt like empathetic and natural.

The Impact 

The outcome of the voice bot integration has been phenomenal for both the business of Meesho and the users. As the bot has to respond to most of the incoming requests, the response time for issue resolution has decreased by ~ 50%. Hence, customers receive quicker responses, without having to wait too long. On the customer experience side, Meesho registered a remarkable 10% increase in customer satisfaction, as customers were happy about the timely and painless support, particularly in their native language. On the business side, the transition to voice-based auto-support also proved to be a massive leap in success, which enabled Meesho to reduce its customer support call costs by almost 75%, enabling it to release its capital to invest in other business areas of the company.

What’s Next: Future AI Features in the Pipeline

Meesho isn’t just considering quick assistance — it has the potential to bring a whole world of convenience and smarts through AI. The platform can be engineered to adapt to provide a more personalized and useful experience for both buyers and sellers.

• Smart Seller Assistants: Meesho can have Smart Seller Assistants that guide sellers to make improved choices, such as recommending the most profitable prices to list, monitoring stock levels, or recommending expanding products that would have a higher probability of selling.
• Emotion-Aware Support: Subsequent releases of the voice bot can pick up on the caller’s mood — irritated, bewildered, or elated — and modulate the tone and content to get closer to human-like and empathetic conversations.
• Personalized Voice Shopping: As voice technology goes deeper in the system, Meesho can actually suggest products to a customer based on their purchase history, interests, and affordability from a voice interface.

These new capabilities can simplify the entire experience of buying and selling by making it easier, smarter, and personalized.

Conclusion

By introducing an AI-driven voice bot, Meesho has made a leap and a half towards the digital-era customer support revolution. It now offers quicker, more local, and cost-for-value service that ideally complements its vision to empower India’s next billion online entrepreneurs. By achieving 90% less response time and operating cost savings, Meesho has created a new standard for leveraging AI to offer a better customer experience in social commerce.

External Resources:

1. Netflix Tech Blog – “Foundation Model for Personalized Recommendation”
Use as:
See how Netflix engineers unified ML models for personalization at Netflix’s official tech blog.

2. The Verge – “Netflix is testing a new OpenAI‑powered search”
Use as:
Check Netflix’s new conversational search set-up here.

~DailyAIWire