Question: I'm looking for a recommendation engine that can handle large amounts of unstructured data and provide real-time results.

If you need a recommendation engine that can ingest lots of unstructured data and provide results in real time, Recombee is a good option. The company offers a powerful AI-powered real-time recommender service with a RESTful API and SDKs. It can handle more than a billion recommendations per day and lets customers add custom filters and boosters with the ReQL query language. It fuses deep-learning, collaborative filtering and content-based filtering for a very personalized approach.

Another good option is Vespa, which offers a unified search engine and vector database that can perform vector search, lexical search and structured data search. It's geared for developers who want to build search applications that scale and perform well, so it's good for recommendation and personalization work. Vespa supports a variety of machine learning tools and has auto-elastic data management for high end-to-end performance and low latency.

If you want real-time adaptation based on behavioral signals, Shaped offers a customizable recommendations and search service that plugs into existing data sources for fast deployment. It can handle multi-modal unstructured data and offers customizable ranking and retrieval components, so it's good for marketplaces, social media, media sites and e-commerce sites. Shaped also offers enterprise-grade security and scalability, as well as white-glove support.

Last, Elastic is a broad Search AI Platform that speeds up time to insight and builds personalized experiences by combining search and artificial intelligence. It offers a range of options, including vector search and generative AI, and is used in industries like the public sector, financial services and retail. Elastic offers a lot of resources, community support and training, so it's a good option for improving user experience and search.