How CyberArk uses Apache Iceberg and Amazon Bedrock to deliver up to 4x support productivity

This article describes how CyberArk achieved up to 95% reduction in case resolution time by combining Apache Iceberg and Amazon Bedrock to automate support operations.

• Legacy system required days of manual data preparation before support engineers could investigate cases
• Amazon Bedrock automatically generates grok patterns for parsing diverse log formats in real-time
• PyIceberg enables single-stage serverless log processing without AWS Glue crawlers
• Iceberg's metadata layer provides instant queryability and partition pruning for faster queries
• AI Agents autonomously perform root cause analysis by querying logs and knowledge bases
• Support engineers now resolve 8-12 cases daily versus 2-3 previously (4x productivity gain)
• Simple cases reduced from 4-6 hours to 15-30 minutes; complex cases from 15 days to 2-4 hours
• Zero manual intervention required for new log formats or schema changes
• Serverless architecture scales automatically with support volume without idle costs

CyberArk transformed support operations by automating data preparation and investigation workflows, enabling faster customer issue resolution and significant productivity improvements.