Meesho is largely dependent on artificial intelligence to propel the next stage of e-commerce expansion in India, from voice shopping agents and vernacular address mapping to AI-generated code and automated customer care.
As the SoftBank-backed e-commerce platform increases investments in AI-led commerce infrastructure, Meesho is quickly integrating artificial intelligence into its operations. Chief executive Vidit Aatrey revealed that over 70% of the company’s code is now AI-generated.
“We have consciously wagered on AI to serve as our operating system. In Meesho’s annual shareholder letter, which was published alongside the company’s Q4 FY26 statistics, Aatrey stated, “We are releasing products faster and with greater reliability than at any point in our history, and over 70% of our code is now AI-generated.”
According to the corporation, AI is now incorporated into every stage of the software development lifecycle, including the creation of code and test cases, code reviews, deployment fixes, and production monitoring.
Meesho’s platform experiments in the March quarter increased from a year prior, according to Aatrey, who added, “What once took weeks now takes days.”
The remarks coincide with a growing number of Indian online companies promoting AI as a fundamental operating layer rather than merely a tool for productivity. However, Meesho seems to be more concerned with speeding up product development and resolving infrastructure issues specific to India’s e-commerce business than with cutting headcount right now.
Dhiresh Bansal, chief financial officer of Meesho, told Moneycontrol, “We are not looking at it right now as a cost lever.” The amount of engineering bandwidth we had to implement specific concepts used to be a major limitation. Now that AI is producing a lot more code, we can accomplish that.
According to Bansal, the company anticipates that AI will greatly increase the speed at which it distributes tests and products around the platform.
Betting big on voice commerce
Conversational and voice-led commerce is a key component of Meesho’s AI initiative, which the business believes has the potential to become the primary shopping interface for India’s next generation of internet users.
According to the shareholder letter, Vaani, the company’s voice AI shopping assistant, reached 1.5 million users in its first month of operation. According to Meesho, conversion rates increased by 22% for users who embraced the feature.
Through genuine voice chats in colloquial languages, the assistant enables users to search, explore, and make purchases.
According to CFO Bansal, consumers who previously had trouble with conventional search-led interfaces are showing significant early traction for the company.
Vaani, our voice AI shopping agent, is one example that we also mentioned in the shareholder letter. Compared to the baseline we had earlier, that has seen an increase of almost 22 percent,” he stated. “Because the interface finally speaks their language, customers who had previously given up without placing a single order are now completing purchases.”
AI tackling India-specific e-commerce problems
Meesho is utilizing AI to address operational issues related to India’s disjointed logistics and commerce ecosystem in addition to shopping interfaces.
GeoIndia LLM, a proprietary AI model trained on delivery traces spanning thousands of pin codes that aids in the interpretation of landmark-based and vernacular addresses—a persistent issue in Indian logistics—is one of its main projects.
The company claims that GeoIndia decreased misroute-related expenses by 5% and increased geocoding accuracy by 20 percentage points.
Additionally, the business is using AI models for customer service, logistical planning, and fraud detection.
Its ML-based network intelligence technology aids in optimizing delivery routing and truck allocation, while its TrustMesh technology forecasts return-to-origin behavior and flags questionable transactions.
According to the company, 62% of customer inquiries are now resolved by Chorus, Meesho’s AI-powered customer care platform, without the need for human participation. Meesho stated that in FY26, the technology contributed to a 23 percent reduction in customer service expenses.
According to the shareholder letter, the company’s AI-driven recommendation engine, PRISM, now drives over 75% of platform orders through intent-driven product discovery and personalized feeds.
AI becoming core to Meesho’s operating model
According to Meesho’s management, the company increasingly views AI as essential to how it develops goods, boosts productivity, and increases access to e-commerce throughout India.
According to Bansal, AI is assisting in the removal of engineering bottlenecks that previously restricted the company’s ability to implement new concepts and introduce products.
In the past, there were enough ideas that needed to be implemented, but there wasn’t enough engineering capacity. He remarked, “We could never carry out all of them as quickly as we would have liked to.” “We can now solve a lot more issues much more quickly.”
Additionally, the corporation is investing in what it refers to as a “autonomous SDLC,” in which software development operations such as coding, testing, review, and deployment are increasingly handled by AI agents.
Meesho believes that economics and AI are strongly related. According to the company, BharatMLStack, its in-house AI infrastructure stack, now operates at 60–70% cheaper inference and AI workload expenses than comparable cloud offerings.
Meesho seems to be portraying AI not only as an efficiency tool but also as the fundamental infrastructural layer driving the next stage of Indian e-commerce growth as it expands into smaller cities and vernacular markets.
