How we used semantic search to make our search 10x smarter

What this post added

This post details Tokopedia's implementation of semantic search using Milvus. It covers their transition from keyword-based search with Elasticsearch to vector representation and similarity search. The decision to use Milvus was based on its ease of use and comprehensive index support compared to FAISS. The post also describes the deployment of a Milvus cluster with Mishards for high availability, and explains the fundamental concept of vector indexing for accelerating similarity search.