Dec 2023 – Present · 2 yrs 3 mos

• Key Achievements & Responsibilities:
• Architected and implemented a real-time data ingestion orchestration system that unified multiple ingestion processes, significantly improving operational efficiency and scalability
• Led a major architecture improvement initiative that resulted in:
• 5.2X improvement in throughput
• 35% reduction in ElasticSearch calls
• Reduced SEV2 incidents from 67 to 0 during peak seasons
• Zero incidents during Prime Days, Black Friday, and Cyber Monday
• 2X increase in system load capacity
• Transformed operational reliability metrics:
• Eliminated daily SEV2 incidents in Q4 (previous: ~30 incidents/month)
• Reduced system downtime from 4 hours to zero during peak traffic
• Achieved 100% uptime for NA and EU regions during critical business periods
• Spearheaded performance optimization projects achieving:
• 90K TPS throughput
• 7-8K searches per minute per region
• Zero downtime during peak business periods
• Eliminated 30-60 minute downtimes previously experienced
• Technical ownership of mission-critical services:
• Led JDK17 migration of core libraries
• Resolved complex infrastructure challenges in VPC deployments
• Identified and helped resolve AWS OpenSearch bugs, benefiting the broader AWS community
• System Reliability & On-Call Excellence:
• Handled highest ticket resolution in team: 128 incidents (45 high severity)
• Maintained 99.99% service uptime during peak business periods
• Implemented robust monitoring and alert systems

Jul 2022 – Dec 2023 · 1 yr 5 mos

• Re-architected the Realtime Ingestion Service to facilitate MultiChannel Downstream Publishing, incorporating a robust Retry Strategy, ensuring zero data loss. Successfully scaled the service to ingest events at an impressive rate of 80K transactions per second (TPS).
• Designed and implemented the UsageMetrics Service from the ground up to capture usage metrics for various client-facing services in both realtime and non-realtime scenarios. Achieved scalability to capture 150 million events per day.
• Developed a solution to identify and flag 44 million duplicate events per day within the Realtime BigData Analytics product.
• Pioneered the design of a Polyglot query engine for the Realtime BigData Analytics product, enabling the querying of 37 billion rows in just 1.1 seconds.
• Successfully launched multiple realtime visibility products, collectively contributing to an impactful 50 million USD.
• Re-architected and implemented a FAN-OUT based deployment pipeline, reducing deployment time by 90% and providing more granular control over service deployment.
• Designed and implemented a MLOPS pipeline for Amazon's Realtime BigData Analytics product suite. Additionally, designed and implemented a data pipeline capable of processing 30TB of data per hour with a 90% compression ratio for ML Model Training, Validation, and Refreshment.
• Consistently contributed to enhancing the resiliency and scalability of the Real-time BigData visibility architecture, managing the processing of 20 billion events per day and facilitating realtime joining of multiple streams in under 1 minute.
• Contributed to team's Operational Excellence by proactively identifying redundant issues during the 2022 Peak. Devised a solution to accelerate the recovery of the ES cluster by 30%. Carried out 4+ POC projects which enhances the current customer experience by re-shaping the realtime analytics.

Jan 2022 – Jun 2022 · 5 mos

• Designed Model Life-Cycle Management Framework to manage life-cycle of ML Models at scale.
• Implemented Real-Time Intelligence Analytics enrichment layer for Amazon’s Global Real-Time Transportation Visibility Engine.