Untitled

Use Case Brief: We are leveraging a GenAI solution to generate SQL queries based on user prompts. These queries are executed on a Postgres SQL database (using Aurora DB Service on AWS) to provide responses to end users.

Problem Statement: The query response time is excessively high, taking 5-6 minutes for one of our data sources, even for simple queries. This is based on testing with only 2 months of data, whereas the production requirement is for 3 years of data. The data source has 40 million records per month. We have tried optimizing it with indexing, which reduced the query time to 2-3 minutes. However, the expectation is for the response time to be under a minute with the complete data volume.