Traditional search engines match keywords. If you search for "dog shelters around Gurgaon" and the indexed page says "animal shelters near Delhi," you get no results. The words do not overlap. Semantic search fixes this by converting text into vectors. Similar ideas end up close together in vector space, even when the words differ. From words to vectors An embedding model takes a word or sentence and produces a high-dimensional vector. The key property: semantically similar inputs produce vectors that are close to each other. "Dog" and "animal" sit near each other. "Dog" and "car" do not. For a search engine, the pipeline is straightforward: Convert every document in the corpus into a vector and store it. Convert the user's query into a vector using the same model. Find the documents whose vectors are closest to the query vector. The hard part is step 3. A corpus of a million documents with 768-dimensional vectors is a massive dataset.…