As a Principal Engineer for AI & Data Platforms, I architect the foundational infrastructure powering next-generation security products. I lead the strategic evolution of our model serving and data ingestion ecosystems, bridging the gap between high-performance systems and rapid AI innovation.

Strategic AI Infrastructure

LLM Serving Platform: Spearheaded the end-to-end architecture of an enterprise Generative AI platform. Standardized internal workflows on vLLM and Triton Inference Server, achieving a 90% reduction in deployment lead time via GitOps-driven automation.

AI Security & Guardrails: Directed the design of a production Prompt Injection Detection System. Engineered real-time security guardrails and a ground-truth validation framework integrated with BigQuery to secure enterprise-scale AI interactions.

Data Platform & Global Scale

Strata Logging Service: Architected the cloud consumer layer to process billions of daily events across 20+ global regions. Maintained 99.9%+ uptime through custom Java/Netty ingestion pipelines and automated Regional Boundary (RBI) deployment strategies.

Security Frameworks: Authored a unified enterprise RBAC framework and optimized AWS Kinesis integrations using a custom AssumeRole strategy, resolving complex multi-tenant security bottlenecks.

Engineering Excellence & Leadership

Velocity Multiplier: Successfully demonstrated a 5x productivity surge in 2025, delivering foundational AI infrastructure (3,000+ commits) in under 9 months.

Modernization: Orchestrated the migration of 400+ Python modules and a full build-system transition to Maven, reducing CI/CD execution times by 60%.

Golden Paths: Established DevOps and GitOps "Golden Paths," reducing new service onboarding from weeks to hours through reusable Helm templates and automated validation.

I served as a lead architect for eBay’s global data ecosystem, focusing on transitioning legacy big-data workloads into modern, cloud-native environments.

Key Achievements:

Cloud Transformation: Led the re-platforming of our massive ETL stack onto Google Kubernetes Engine (GKE), which improved service recovery times from 30 minutes down to just 2 seconds.

Strategic Migration: Engineered "Dual-Run" frameworks that allowed us to migrate mission-critical production data from Teradata to Spark/Hadoop with zero downtime and 100% data integrity.

Platform Democratization: Architected a generic ETL methodology that empowered thousands of internal power users to run 100K+ daily production jobs across diverse data environments (Oracle, Teradata, HDFS).

Managed one of the industry's largest Hadoop clusters (2,000+ nodes, 10PB+). My role involved deep-core performance tuning of HDFS, MapReduce, Yarn, and HBase to support high-intensity data science and analytics workloads for over 2,000 active users.

At Microsoft, I worked within the Windows Phone division, focusing on the intersection of UI performance and core OS connectivity. I engineered system-level features for WiFi and Bluetooth stacks, ensuring seamless hardware-to-software integration and high-performance IPC.

Developed high-performance cellular infrastructure and network management protocols. This role provided my foundation in high-concurrency C++ development, low-level IPC mechanisms, and building resilient systems for global telecommunications networks.