Sep 2024 – Present · 1 yr 9 mos

• Promoted to SDE IV, currently leading key initiatives in the Metrics Store team and driving business-critical improvements across performance, resilience, and platform modernization.
• Led Kubernetes cluster monitoring enhancements by designing and evaluating next-generation Metric Store technologies. Delivered POCs with HBase and VictoriaMetrics, and drove the GA rollout with Splunk Metric Store.
• Designed generic entity and metric APIs with extensible pagination and Redis caching to support broad AppDynamics use cases.
• Upgraded the open-source Java client with a high-throughput async framework, enabling large-scale metric ingestion with a minimal thread pool.
• Identified and fixed issues in the Splunk SignalFx WebSocket library, resolving connection handling and resource blocking problems through patched implementations.
• Led remediation of high-priority security vulnerabilities, improving deployment velocity and platform stability.
• Modernized the platform by migrating a large multi-module Gradle codebase from JDK 8 → JDK 17, upgrading to Dropwizard 3.x, and revamping the test framework for faster developer productivity.
• Reduced reported-nodes job runtime by 10× through concurrency and pipeline optimizations.
• Tuned Java performance using G1GC and ZGC, reducing GC pauses and improving overall latency.
• Delivered an HBase storage-optimization POC, achieving 6–10% storage savings and improving batch-write and scheduled-job throughput.
• Optimized Spark jobs processing terabyte-scale data, reducing execution time by 24% and lowering CPU consumption.
• Scaled metric ingestion services to reliably process large, concurrent payloads in high-throughput environments.
• Improved HBase resiliency, mitigated hotspotting, and tuned CMS GC on DB nodes to enhance cluster stability.

Oct 2022 – Aug 2024 · 1 yr 10 mos

• Part of the Data Platform team ingesting billions of metrics/min, 100+ TB/day, and managing petabyte-scale data. Key contributor to performance, cost optimization, and platform resilience.
• Added rollup validation features to detect metric aggregation corruption across tiers—helped renew multi-million USD customer deals.
• Improved Reader-Writer service performance and migrated from CMS to G1GC across 30+ clusters (handling up to 85M metrics/min), achieving 25% cost savings and a $400K/year reduction in compute costs.
• Tuned Dropwizard apps, HBase reads/writes, and performed EC2 cost analysis to maximize performance efficiency.
• Introduced enhancements like custom alert visibility, connection pooling, exponential backoff, and smarter error handling in Reader services.
• Optimized Java-based HBase migration jobs and contributed to reducing on-call alert noise, improving operational stability.
• Handled critical P0–P1 escalations, becoming the go-to team contact.
• Simplified local setups and conducted sessions on IntelliJ-based debugging and developer productivity.
• Authored and improved 45+ technical documents to enable easier onboarding and system understanding.
• Led AWS EMR POC for HBase migrations: monitoring, failover routing (multi-AZ), tuning, and performance testing.
• Resolved ~80% of critical-to-medium codebase vulnerabilities (BlackDuck) to enhance security posture.
• Researched and evaluated multiple DBs for building a next-gen time-series Metric Store.