Working on regionalizing components of Gmail and Chat architecture with respect to at rest and processing data. https://workspace.google.com/blog/product-announcements/announcing-sovereign-controls-for-google-workspace

Conceptualized and implemented an architecture to support private network access for all end user touchpoints in the product. The architecture restricts all network traffic between customers and our service within Oracle Cloud Infrastructure. Developed a solution that allows customers to access inbound service endpoints only from within a private subnet provisioned in their Oracle Cloud Account. Additionally, it supports outbound connectivity from data plane components to customer resources. Spearheaded the technical team and worked with Dataflow and DNS teams to support ETL and transparent zone features.

Designed and developed an Application Gateway Module to expand Fusion Analytics product’s offering to NetSuite Application Customers. Abstracted out Identity, Database and Cloud Operations into a single simplified integrated REST API that could be used by consumers for managing Product Lifecycle. Integrated the Gateway with different microservices for operational management like Data Warehouse, DNS Service, Identity, Logging and Telemetry.Delivered the beta phase to production in less than 6 months and worked continuously with Product Managers and NetSuite teams/Beta Customers to enhance features and ease of use. The offering went GA with close to 20 Enterprise Customers and 50+ in beta stage.

Designed a state machine that optimizes resource consumption of DHCP Relay based on interface and server configurations.Implemented a command line plugin in Arista EOS which can dynamically modify the lease table based on options configured.Designed and implemented a component of the synchronization protocol across a distributed operating system.

Built a scalable and secure IOT hub which acts as a gateway for routing data to AWS S3 storage component from an android smart watch.Implemented a double level encryption and authentication scheme to make the architecture robust to security threats.

Worked as a Graduate Teaching Assistant for the course LS30B – Mathematics for Life Scientists.Responsible for conducting two lab sections, grading assignments & exams papers and holding office hours

Created a browser extension for Chromium that is capable of classifying webpages into distinct categories. The extension was developed in python and deployed on the Tizen Operating System.Used Document Object Model Parser to scrape relevant data from over hundred thousand URLs which was fed to a series of classification models including logistics regression, N-Nearest Neighbor Search and Bayesian Network - some of which required the use of data compression techniques. Selected Bayesian Model as the most optimal and accurate model for prediction.Used Ada-boost on different models to achieve an accuracy of 90%.

Created a dynamic website which provides a list of available restaurants according to the location selected. Implemented k-Nearest Neighbors to classify a new restaurant.