• Develop system to automatically deploy hybrid cloud infrastructure to run 3rd party enterprise applications on Google Cloud Platform.• Serve as tech-lead for the business continuity and disaster recovery domain, automating backup and restore processes and developing a long-term roadmap for disaster preparedness.• Automated backup creation for all new services on Google Kubernetes Engine using Terraform to ensure recovery point objective of 1 hour and remove 1-2 days of developer time.• Coordinating with multiple teams and work streams to lead multi-year project that will improve service reliability and reduce developer toil by adding integration and end-to-end tests for our service.• Automated DNS and access control list configuration for services deployed on Google Kubernetes Engine.• Created a POC for an LLM to auto-diagnose alerts based on existing playbooks and make recommendations for how to fix them using a vector database and chunking the playbooks into embeddings to find relevant playbooks.• Improved the security posture of 100s of services by migrating to a new group management system that enforced two party approvals and removed unilateral access to robot users.• Defined SLOs for our service to ensure a good customer experience and implemented observability solution to help us detect issues early to maintain our SLO.

• Enabled the business intelligence team to measure KPIs to drive future business decisions by designing and developing a poller that captures updates from a graph database and transforms them into a relational format, written in Java.• Implemented caching in latency sensitive APIs which reduced the latency by 50% for cached responses using Java and AWS Cloud Formation.• Wrote and designed APIs to support account configuration in Prime Video Direct using Java.• Ensured the operational success of services by adding integration and load tests, metric monitoring, one box hosts, and automated code coverage measurements using Java.

Developed a web app to deploy a machine learning based forest fire prediction model using react and flask.Performed data preprocessing and cleaning on a large car crash dataset

• Assisted a class of over 100 students in concepts such as web development, basic networking theory, multithreading, and databases • Lead labs focused on practicing these skills • Held weekly office hours designed to help undergraduate students with topics covered in class

Updated the Alaskan DOT android app, Alaska 511 to API level 26.Fixed various bugs that were causing the app to either crash or not behave as expectedDeployed update to google play store where it’s been downloaded by thousands of users

• Worked part time with Boston College, Lesley University, and Williams College Spanish literature professors to construct a website for a Spanish literature review journal• Built website using WordPress, Divi theme, and custom code

• Developed a recommendation engine for a simulator to analyze results of a simulation and provide feedback for how it could be better optimized• Incorporated stochastic analysis of individual recommendations to give recommendations that worked even with variations in the simulation 
• Performed early stage research to develop a multi-view skeleton tracking system using RGB-D cameras through the use of deep learning computer vision algorithms

• Performed feature extraction on EEG data of people performing everyday tasks• Utilized extracted features to train an SVM to classify different EEG wave patternsDeveloped live collaborative whiteboard using react and meteor