AI-powered performance recommendations for Amazon Redshift
Big Data Blog
This article demonstrates how to build an AI-powered solution using Amazon Bedrock and Lambda to automatically analyze Amazon Redshift performance telemetry and deliver prioritized optimization recommendations.
- Collector Lambda runs diagnostic SQL queries and CloudWatch metrics, pre-computes performance signals, and writes telemetry to S3
- Analyzer Lambda reads telemetry, builds structured prompts with signal-to-metric correlations, and calls Amazon Bedrock Claude for recommendations
- Solution detects table-level issues like row skew, stale statistics, and sub-optimal keys, plus runtime problems like disk spill and small-insert bursts
- Recommendations include priority levels, categories, signal sources, explanations, SQL actions, and expected impact estimates
- EventBridge triggers daily 24-hour runs; SNS sends email summaries of top 10 recommendations to subscribers
- Signal-based design produces specific, workload-targeted advice rather than generic best-practice guidance
The solution enables automated daily performance reviews for Amazon Redshift Serverless with minimal operational overhead and cost-effective Amazon Bedrock usage.
The AWS News Feed is currently looking for gold sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.
Related articles
2026
2024
2026
2026
The AWS News Feed is currently looking for silver sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.