Need and complexity of analytics
Modern applications rely heavily on low-latency, high-throughput access characteristic of OLTP systems. Users, whether human or other applications, expect immediate responses from client applications or web interfaces. No customer ever says, please slow down the performance; it is too fast for our end users, and we want you to take more money for database services while you are at it.
DynamoDB excels at OLTP, making it an ideal choice for such workloads. However, data needs to be accessible not only to end users but also to internal business and data science teams. These teams require the same data for various analyses, demanding different formats and access methods to run algorithms that generate actionable insights. This necessity drives the need for analytics.
Technically, one could perform a full table scan on a DynamoDB table to retrieve data for reporting or machine learning pipelines. This naive approach, however, is impractical for large...
