Developed an asynchronous C++ HTTP server which processed requests received from Instagram's C++ process manager

Bridged communication between the C++ server and Python Django using a Cython module

Invoked Django request processing to return a response to the C++ server

Rearchitected Android Real- Time Communication messaging infrastructure to improve performance and future deployments

Modularized overcomplicated Java and Kotlin classes to simplify dependencies between Buck modules

Analyzed an ML model used to identify high severity data loss issues from Windows user's feedback

Processed feedback data from over 12 million monthly device from Windows Feedback Hub in Azure Data Explorer from an Elastic Search database using Kusto Query Language

Adapted the internal Python ML model to run locally in Jupyter Notebook using Apache Spark,

numpy, and pandas before using Sci-Kit Learn to analyze performance and generate decision trees

Developed a data insight tool which provides performance metrics to Windows developers in Microsoft's Visual Studio/Code editor

Processed around 10 TB of raw data from Microsoft's internal NoSQL Azure Cosmos DB and used Azure Data Factory to transfer cleaned data into Microsoft SQL Server Management Studio

Created a user interface using C#, XAML, and Markdown that provides a popup GUI in Visual

Studio/Code that displays insight information when hovering on a line of code