Edward Greenberg

• Tech Lead since 2022

• Created tooling for indexing using YAKE! keyword extraction, with HuggingFace, Gradio, and PyPDF2 • Ongoing research on other machine learning applications for academia

• Created tools for (1) generating sequences, (2) finding mode centrality, and (3) mocking data, strengthening points of interest in NPM’s ecosystem • Navigated NPM development environment to ensure deployment and package format compatibility with other host page and other installations • Tested packages in local Node environments, increasing promise of user-facing reliability while verifying outputs/laying new version groundwork

• In React, utilized Flux architecture, the virtual DOM and unidirectional data flow to create a cleaner environment for debugging, and to reduce code duplication through leveraging modularity of functional and class components to create a dynamic JSX SPA, easing code refactoring • Leveraged WebSocket to ensure compatibility between devices and real-time user communication, providing persistent and secure connections between client and server by using bidirectional security models and long-living asynchronous messaging, reducing need for server polling • Implemented Webpack configuration to bundle modularized file structure/essential static assets with Babel transpiling, converting ES6 and JSX to browser JS, enabling hot module replacement via proxy server to aid in development/debugging, and uglifying/minifying code to optimize render

• Utilized React, Redux, and React Hooks to manage views, centralize state storage, and access the capabilities of different lifecycle methods • Worked with Node.js, CLI tooling, ES6 syntax, and Git to create a scalable 2D digital environment capable of handling a variety of user inputs, of offering up dynamic events, and of centrally storing large amounts of data parsed by complex bespoke algorithms while traversing different views

• Developed and maintained high-priority business data wrapper software with Node.js, GraphQL, React, AWS, Python, and other fullstack technologies • Built ‘smart’ upload tool dynamically reading spreadsheet fields and allowing for immediate DynamoDB memorization of novel field titles • Reduced overlapping code in one ETL pipeline by over 2,000 lines by leveraging design patterns and genericization principles--this pipeline moved data from a React frontend, through various filters and checks, through a GraphQL / Lambda Node.js orchestration layer, and to Postgres

• Utilized Typescript to improve code readability through strict type annotation while creating an easy platform for compiling, making code compatible with a broad variety of old and new JavaScript configurations, all of which frontloaded error detection, and reduced cognitive overhead in code integration • Implemented GraphQL schemas and architecture to ensure product compatibility with APIs and leverage native GraphQL user feedback with error handling, using queries to create scalable requests fetching precise server data reducing response payload in the context of a schema hierarchy • Utilized Express framework and structured server environment to connect primary project code to SQL and non-relational data, writing request handlers for different HTTP methods to direct to different URL paths or routes, exposing a dynamic and lightweight API for controller logic • Navigated VS Code development environment to ensure marketplace release and compatibility with Microsoft Azure architecture, using tools such as version instancing and linting to remove upload conflicts, maximize usability, and modularize application to connect local to user file structures • Implemented Mocha/Chai testing framework with Istanbul/nyc coverage display tools to ensure user-facing reliability and increase iteration speed when adding new applications features, enabling incremental test-driven development complementing a scrum model clarifying programmer roles • Managed user file structure by developing automatic bash CLI commands to increase application ease of use and portability, and connected these CLI commands to a robust series of parsing algorithms enabling automatic queries and config-filling while responding to user save commands • Product developed and maintained as part of OSLabs

• Leveraged Python engine to assess and deploy analytics against neuron-fire outputs, using Python’s increased human semantics relative to other languages and robust data science library to clean and process input, and developed a new semantic database of findings using network analysis • Used Arduino to handle locally-produced engineering solutions, maintaining and troubleshooting atypical usage of a treadmill with Qualisys motion capture for the purpose of introducing a somatic component to neural analytics, all in the context of research, product development, and trainings • Deployed lab-created brain-computer interface (BCI) technology to assess, understand, and modify machine-neural event loops (“mind-reading”) with live error detection and robust feedback cycles; acted as tech evangelist, developing BCI applications for users with potential motor disorders

• Managed relational data with R, Stata, VBA, and SPSS; led large research project on violent behavior in Poland, Ghana, and the United States • Taught Statistics, Research Methods, Ethics, and similar classes at the college level, working with databases to manage grading/course content

Worked as a reporter, photographer, copy editor, and interim news desk manager for Albany’s chief political newspaper; interviewed State legislators, politicians, scientists, and political activists; covered press conferences led by Governor Andrew Cuomo, and U.S. Senator Kirsten Gillibrand