Ajay Sagar Bokka

• Built and launched the CloudTrail Insights feature in 5+ AWS regions, performing rigorous testing and validation, leading to a 60% increase in customer adoption.
• Designed and implemented real-time event processing pipelines using Apache Flink, Amazon Kinesis, and DynamoDB, ensuring scalable, fault-tolerant, and deterministic data processing.
• Automated on-call operational tasks, reducing manual interventions, and cutting engineering effort by 40%, leading to a 15% decrease in incident response time.
• Addressed critical memory usage spikes in high-volume regions, which led to a daytime severity level 2 incident, posing a risk of memory exceeding 100%, potentially dropping attribution data, and impacting insights delivery. I took immediate action by increasing the severity of the ticket to accelerate resolution. Collaborated with the team to reassess shard allocation, considering prior increases in high-volume regions, and successfully stabilized memory usage to prevent data loss.
• Upgraded all components to Java 17, significantly enhancing efficiency and optimizing memory usage across regions. This migration improved HeapSpace and MemoryUsage, ensuring better resource utilization, and system stability. Successfully delivered a seamless transition with enhanced performance.
• Automated the handling of production resources during large-scale events by onboarding multiple Python and Bash scripts, ensuring system safety, and reducing manual intervention effort by 40%.
• Enhanced Model Configuration with density information to improve insight compaction in the SSF algorithm for ACR Dense. Modified key components to pass DatasetType information, and refine the compaction logic, ensuring more accurate insight generation and processing.
• Acted as Scrum Master, leading sprint planning, coordinating with multiple teams, and addressing on-call incidents to maintain service reliability.

Account-Level Backfilling System Enhancement.

• Implemented message handling components to differentiate between API-level and account-level backfill requests, ensuring a smooth transition during the system upgrade.
• Engineered test cases and performed load testing to validate system behavior under various load conditions.
• Integrated monitoring solutions to track key metrics, including: message processing rates, backfill completion times, error rates, and system health indicators.
• Extended the existing codebase to support new message formats while maintaining backward compatibility.
  

Real-Time Event Processing with Apache Flink and Amazon Kinesis.

• Developed a real-time data pipeline using Apache Flink to process millions of events from Amazon Kinesis.
• Implemented event-time processing with tumbling windows for correct ordering and late-event handling.
• Applied distributed staggering to prevent simultaneous window closures, reducing peak computational load.
• Used Amazon DynamoDB for storage, and optimized batch writes for cost efficiency.
• Ensured fault-tolerant processing using Flink checkpointing and recovery mechanisms.

Region Expansion:

• Manually built and launched the CloudTrail Insight feature in 5+ AWS Regions.
• Performed infrastructure deployments and built components with modified system configuration as per the region's guidelines.
• Validated the Insights Workflow and public APIs using Bash scripts.
• Automated the region expansion workflow to reduce engineers' effort by 40%.