Enhancing and optimising large-scale Machine Learning / data processing pipelines across Google

Designed and taught graduate-level courses on Distributed Systems, Data Processing and Pipelines, and Artificial Intelligence systems.

• Led the development and deployment of a stacked ensemble ML model for forecasting EV charging demand, reducing error by ~55%, and optimized data storage using a custom MongoDB database for charging profiles.

• Designed and developed algorithms (using C++/Go) for a scalable EV smart-charge scheduling system, improving grid system congestion and power constraints by ~25%

Reduced downtimes by ~95% for a database introspection service through:

o Developing state persistence through a distributed filesystem-based checkpointing and recovery mechanism

o Using a leader-election algorithm to enhance availability by switching between task-replicas.

Selected as a candidate for IDEA Global Internship 2023

▪ Developed microservices (using Spring Framework) and APIs for improving peak systems capacity by ~15%

▪ Coded and maintained features, improved performance, and bolstered security for web applications through

back-end development using Node.JS / PHP

▪ Designed, implemented, and scaled MySQL databases

▪ Enhanced event logging and alerting capabilities (Prometheus and OpenTelemetry) for applications and reduced

system downtimes by over 25%

▪ Tested, validated, and automated deployments using Jenkins to reduce deployment times by ~30%