Zalak Bhojani

I have led and delivered high-impact initiatives across AWS EMR Serverless, Glue, and SageMaker, owning architecture and execution across control plane and data plane components. Known for driving complex, cross-team programs to build distributed, scalable, and secure cloud platforms. • Led a 5-engineer team to deliver continuous scaling for Slurm-orchestrated SageMaker HyperPod clusters, enabling dynamic capacity management for large-scale ML training workloads previously limited to EKS • Improved operational excellence by strengthening code review practices, streamlining customer issue handling, and promoting consistent knowledge sharing across the team • Led and mentored junior engineers to bring Glue Interactive Session to GA for a regional launch, using GenAI-driven automation to enable a one-touch release

Contributing to the control plane of AWS EMR Serverless, enabling scalable, serverless big data processing. • Designed a real-time resource utilization system that exposes per-job metrics to customers, scaling to support 20,000+ concurrent jobs. Event-based updates via EventBridge enabled other AWS services to build metering and billing workflows • Built customer-facing APIs for CloudWatch logging, improving observability and debugging capability for serverless Spark workloads • Optimized disk usage and ECS task runtime by introducing log rotation and compression for on-disk customer logs, reducing job shutdown time by 90% • Designed and implemented public APIs for Spark Streaming workloads, enabling real-time data processing. • Led migration of ~30 microservices to JDK 17, boosting performance and maintainability.

Part of Percy Platform team, Percy is a visual testing product acquired by BrowserStack in 2020. The Percy Platform team owns rendering infrastructure, manages GCP-based cloud infra, and drives integration with BrowserStack’s web and mobile platforms. Key Contributions: • Offered PPO based on strong performance. • Led enhancements for Chrome on Android (Beta): enabled multi-device support per host, high-res image capture via ADB, certificate deployment, and monitoring/alerting setup. • Optimized snapshot performance for large (~50k px) Chrome screenshots — reduced rendering time from ~6.9 mins to 2 secs. • Added emoji rendering support in Chrome and Firefox. • Built a configuration validator to ensure seamless test setup

• Created well documented data structures and algorithms notes which included c++ stl, trees, graphs, dynamic programming,etc for the community.

• Designed, coordinated with other interns of the group and built a web app using flask framework. • Built a Python Tutorial web app, consisting of blogs, videos, user authentication, etc.

CGPA: 8.78/10 (Absolute Grading)