In an online marketplace, users often search using intent rather than exact product keywords, for example, ‘cheap laptop for video editing.’ Traditional keyword search struggles because no product is described exactly that way. Semantic search addresses this by understanding the meaning of the query and matching it to products based on concepts, not just text matches. This makes search results more relevant, but it requires a different approach than traditional keyword-based search, such as using embeddings and machine-learning models rather than simple text matching