Traditional ecommerce search requires shoppers to speak product-database language: category names, color tags, attribute keywords. A shopper who wants "something that feels expensive but isn't" has no filter for that. A shopper looking for "what works with the kurta I bought last month" has no search interface that understands the reference. Both shoppers end up browsing manually or leaving. Natural language search bridges this gap. The shopper expresses their need as they would to a human — any combination of occasion, vibe, constraint, style preference, or existing wardrobe reference — and the system translates it into a catalog query. The output is products that match the meaning of the query, not just the literal keywords. For fashion specifically, natural language search unlocks the full range of how shoppers actually think about clothing: by occasion, by feeling, by what they're trying to achieve, and by reference to things they already own. This is categorically different from keyword matching, which requires the shopper to already know what kind of product they want.