
Storyboard18
Storyboard18
Zepto has redesigned its search functionality to accommodate how Indian users naturally phrase queries, co-founder Kaivalya Vohra announced on LinkedIn. The quick commerce platform recognized that users often search using everyday language rather than formal product terms - typing *"kapde rakhne ki bag"* instead of "laundry basket," for example.
Vohra explained that such searches previously returned no results or irrelevant items on most apps. To address this, Zepto built a system powered by custom large language models that can understand regional language inputs, spelling variations, and user intent.
Industry observers noted that e-commerce search is fundamentally a synonym and intent problem. One commented that high-intent queries often fail not because products are unavailable, but because catalogs are structured around English SKU names rather than how Indian users actually search.
The development reflects Zepto's strategy of designing for Indian user behavior rather than relying on direct translations, as the company prepares for its public market debut amid intensifying quick commerce competition.

PM Modi's Thirukkural Gift Puts Tamil on World Stage

Konkani Language Champion N. Purushothama Mallya Dies at 97

arXivLabs Framework Description Contains No Language Focus

Bodhan AI Launches Four Indic Models for OCR, Speech and Translation

Centre Opposes English as Native Language in CBSE Row

Bangladesh Seeks Bengali-Language AI Models for Classrooms
