Dominic Kramer Email and Phone Number

Tech lead for the Delta Live Tables full stack team

Front-end Codebase Redevelopment===========================* Led a team of 3, developed a transformation tool to migrate JavaScript code to Typescript for several Google’s front-end codebases (Google Hotel Search, Google Image Search, Google Photos, and YouTube) leveraging deep insights into each of their technology stacks and codebases* Designed and implemented a tool to automatically resolve cross-reference issues caused by unsynchronized changes in JavaScript and template code. Reduced development cost by 10%* Successfully completed the redevelopment and deployment of the new code bases with no downtime or outages* Documented the transformation tool and transferred knowledge to front-end development teams.* Acted as owner and point of contact for ongoing collaboration and continuous supportAutomatic Diagnostics Engine======================* Led a team of 2, developed a diagnostics engine to simulate human code generation in Google front-end code in order to automatically analyze the most impactful areas that need improvement. This engine leverages Go and Google Spanner and SQL* Applied MapReduce processing to resolve the automatic code generation barrier related to code bundling across Google’s front-end codebase, which is the major pain point identified by the diagnostics engineAutoML Through Evolutionary Strategies=============================Researched, designed, and in the process of implementing an AutoML approach to automatically learn machine learning architectures using evolutionary strategies with lambda-calculus. This work is a collaboration with Google BrainMentoring=======* Mentored a junior engineer to replace a section of Google’s TypeScript linting stack resulting in an 8x improvement in performance

GCP Node.js Kubernetes User Experience==============================* Invented a patent-pending technique to containerize Node.js applications that does not require Docker locally or in the cloud when deploying Node.js applications to Google Kubernetes Engine or Google Cloud Run* Designed and implemented an end-to-end Node.js utility that integrates with Google Cloud Code and Google Cloud Run. This utility is capable of reducing the difficulty of deploying a Node.js application to Kubernetes or Cloud Run to executing a single commandGoogle App Engine Flex Node.js App Deployments=====================================* Owned the end-to-end developer experience for deploying Node.js applications on the App Engine Flex environment* Designed and implemented the Node.js specific infrastructure that is used every time a Node.js application is deployed on App Engine FlexGoogle Cloud Platform Node.js Diagnostics Beta Availability===========================================* Owned the end-to-end development and launch of the GCP Node.js Cloud Debugger, Logging, and Error Reporting offerings to beta availability* Added support for TypeScript, CoffeeScript, and Windows debugging in the Debugger, expanded support for Bunyan and Winston in the Logging library, and expanding support for Express, Hapi, Koa, and Restify in the Error Reporting libraryAutomated Content Detection in Math Documents through Deep Learning=====================================================* Developed a convolutional neural net in Python TensorFlow to detect, classify, and extract text from theorems, definitions, axioms, and conjectures in mathematics PDF documents to provide training data for a deep learning model used in automated theorem proving. This work is a collaboration with Christian Szegedy’s team in Google Research

Build Infrastructure Developer Experience==============================* Drove the effort to unify all Java (Ant, Grandle, Maven), C++ (make), and Node.js (Gulp, Grunt) build infrastructure across the engineering team to a single approach backed by Artifactory to enable key product feature development* Developed a React web application to analyze data from specialized embedded devicesEmbedded Stream Aggregation=======================* Developed an end-to-end Java solution from scratch to run within an embedded environment to process and aggregate hundreds of data entries per seconds into a unified form to enable downstream processing for a key customer

* Owned and led the development of a custom designed Android system to provide unique functionality, using C++ and Android applications, for several key Boeing customers* Developed and implemented an end-to-end testing framework utilizing the Android Debug Bridge to reduce the time needed to perform quality assurance testing from over an hour to minutes

* Full-stack engineer on a large internal web application powered by JQuery, hand-written JavaScript and CSS, Django, SQL Server, and Apache, used by tens of thousands of employees