This demo uses short descriptions of startups to perform a semantic search. Each startup description converted into a vector using a pre-trained SentenceTransformer model and uploaded to the Qdrant vector search engine. Demo service processes text input with the same model and uses its output to query Qdrant for similar vectors. You can turn neural search on and off to compare the result with regular full-text search.
This demo uses data from Delivery Service for Berlin. Users may like or dislike the photo of a dish, and the app will recommend more similar meals based on how they look. It's also possible to choose to view results from the restaurants within the delivery radius.
E-commerce products categorization
This demo shows how you can use vector search in e-commerce. Enter the name of the product and the application will understand which category it belongs to, based on the multi-language model. The dots represent clusters of products.