I use it for database. For RAG, you need a vector store to store embeddings. To store the vectors, you need embedding models to convert the data into vectors. You then need to store those vectors in any vector store. Popular ones are like Chroma DB. As a new alternative, I selected OpenSearch, which falls under the whole AWS infrastructure. So to bring our full architecture into AWS, I use OpenSearch as a service as my vector store.
Find out what your peers are saying about Amazon Web Services (AWS), Datadog, Dynatrace and others in Application Performance Monitoring (APM) and Observability. Updated: November 2024.
I use it for database. For RAG, you need a vector store to store embeddings. To store the vectors, you need embedding models to convert the data into vectors. You then need to store those vectors in any vector store. Popular ones are like Chroma DB. As a new alternative, I selected OpenSearch, which falls under the whole AWS infrastructure. So to bring our full architecture into AWS, I use OpenSearch as a service as my vector store.
We use the solution as a login platform. We have a lot of microservices, and we get log records from there, which we host on Amazon OpenSearch.