David Sertic

David Sertic, an author, published an insightful article titled ‘Neural Search Tutorial’ on June 10, 2021. In this article, he thoroughly discussed the core concepts of neural search, distinguishing it from regular search methods, and outlined the suitable neural networks for search-related tasks. Sertic provided a comprehensive step-by-step guide on how to build a neural search service using technologies such as BERT, Qdrant, and FastAPI. He explained the process of preparing data for neural search, which involves encoding descriptions into vector representations using pre-trained language models, specifically using the ‘all-MiniLM-L6-v2’ model due to its efficiency in terms of low memory consumption and fast inference. Additionally, he demonstrated the use of Qdrant as a vector search engine to store and manage these vector representations and to perform nearest vector searches. Sertic also showcased an online demo of the neural search service to illustrate its effectiveness in finding semantically similar startups based on their descriptions. Furthermore, he invited readers to join a Discord community dedicated to discussing vector search, similarity learning, and broader applications of neural networks and neural search.