Dive into latest features of Amazon DocumentDB - Lets Talk About Data

In this episode of Let's Talk About Data, host Prasad Matkar is joined by guests Kaarthiik Thota and Inderpreet Singh. The show featured a comprehensive discussion about Amazon DocumentDB, a fully managed document database service from AWS that offers MongoDB compatibility. The session explored how Amazon DocumentDB provides flexible schema capabilities while handling JSON workloads at scale and demonstrated practical applications through a movie search application showcasing different search capabilities, including text search, semantic search via vector embeddings and hybrid search approaches.

Key highlights discussed in the show:

• Native text search functionality allowing searching within text content using MongoDB-compatible dollar text operator

• Vector search capabilities enabling semantic search through integration with Bedrock embedding models

• New NVMe instance types offering improved performance for read-intensive workloads through local SSD storage

• Hybrid search combining both text and semantic search for enhanced search results

• Automatic storage scaling from 10GB to 128TB with pay-for-what-you-use pricing

• MongoDB API compatibility allowing seamless integration with existing tools and applications

• Separation of storage and compute architecture providing better scalability and durability with 6 copies of data

• Best practices for implementing vector search including starting small, testing with sample datasets, and optimizing search parameters

Prasad Matkar - Database Specialist @AWS

Kaarthiik Thota - Senior Specialist Solutions Architect@AWS
Inderpreet Singh - Sr Product Manager@AWS

Links from today's episode