About Me

May 2022 - Present

• Designed a customer contract migration strategy and implemented it in Java, utilizing AWS SNS, SQS, and Lambda to process up to 10,000 contract update requests per customer, allowing customers to perform bulk updates to their titles.
• Stood up two production-ready Java microservices and performed a lift-shift-refactor extraction from a monolith to microservices and enabled full CI/CD deployment pipelines, reducing the average time to production by up to 700%.
• Led the migration of primary customer workflows for audiobook creation from JSP to TypeScript using React.js, resulting in a refreshed experience for 70% of all user flows in the application and improving page load time by 30% serving 80k customers per day.
• Engineered a canary release mechanism in a multi-cell Java microservice architecture, ensuring predictable traffic distribution during cell rebalancing. Reduced on-call pages by 5 per week, improving system reliability and cutting escalation response costs by an estimated $50K/year.
• Implemented a custom backoff strategy in our Rust and Java microservices to mitigate prolonged customer throttling issues when processing large customer requests, reducing on-call pages per week by 10%.

June 2021 - May 2022

• Performed a project rewrite, converted legacy Ruby webapp into enhanced React app using a Model-View-Controller pattern to enhance source code organization
• Optimized the React app and reduced load times by up to 90% compared to legacy Ruby app through asynchronous API calls, better page design, and lazy loading methods
• Improved user satisfaction by iteratively reimagining the webapp UI to be friendlier and optimized for faster workflow, while maintaining core attributes to maintain user muscle memory
• Automated legacy Ruby workflows in Python, eliminating manual file processing and cutting weekly labor costs by $500, saving approximately $26K annually while improving task throughput by 70%.
• Designed and implemented a backend API using Python Flask, nginx, and redis to enable job queuing, and Kubernetes/nginx for load balancing by automatically deploying more pods when traffic became overwhelming.

September 2020 - May 2021

• Implemented several front end features such as a sign-up page using functional, component-based React.js for production
• Redesigned the user interface with Figma to create a friendlier UX/UI based on feedback from professionals in web design
• Streamlined backend integration testing through use of a Docker image
• Maintained an agile workflow through the use of asynchronous task delegation, weekly review meetings, and a task backlog similar to scrum to improve overall efficiency of the development process

January 2024 - May 2024

• Designed a message producer/consumer system to process web socket requests on a Node.js server, and a room system for organizing players into rooms with unique IDs.
• Designed a game state machine and implemented the core gameplay loop using Next.js, NextUI, and functional React.
• Integrated with AWS Bedrock and utilized Llama 2 70B, engineering a prompt to accept user input and generate the appropriate output.
• Implemented a leaderboard system to display the number of round wins each player has at the end of each round.

March 2023 - May 2023

• Designed a web application using React.js, Node.js, and Python that allowed users to create an account and share art generated natively in-app using a self-hosted gen-ai model.
• Self-hosted a generative AI model to reduce operational costs by up to 80% when compared to using OpenAI APIs.
• Utilized AWS SNS, SQS, and Lambda to trigger requests to the server for executing art generation requests.

August 2021 - August 2022

• Designed a Swift application with Node backend that showed popular public events in the users geographic area on a map, and get invited to private events.
• Constructed an intelligent algorithm that can determine the trustworthiness of accounts on the application to assure the safety of event attendees and hosts.
• Built a complex account network using AWS RDS and connected it to the Swift frontend using Node.js and Express.js to enable full CRUD capabilities.

March 2021 - May 2021

• Designed a PacMan machine learning agent that made use of reinforcement learning to play an adversarial game of Capture the Flag against other similar agents
• Created a feature extractor that decomposed state-action spaces into useful features that our agent could attend to, learn from, and weight in order to make optimal moves
• Composed a reward function that provided a balanced set of rewards for positive and negative outcomes of movements made by the agent to encourage meaningful decisions

July 2020 - June 2021

• Published a Discord automation bot built with Python that provided two-factor authentication email sign-in verification features for the official school chat server
• Ensured server security with domain-locking functionality, allowing administrators to force users to use their organization emails when joining the organization Discord server
• Utilized the Gmail and Discord Bot APIs to handle email distribution and asynchronous messaging between the bot and end users

November 2020 - December 2020

• Utilized Python sklearn to perform Logistic Regression on a 47,000-value Pandas DataFrame to predict salaries of individuals based on the real income data
• Observed, sanitized, and stored over 47,000 lines of unorganized and incomplete data from the UCI Adult Income dataset using sklearn's OneHotEncoder and Pandas DataFrames
• Implemented an email spam filter built with the same Logistic Regression model as the Salary Predictor to increase familiarity of sklearn and Pandas libraries, as well as building logistic regression models

August 2020 - May 2022

August 2018 - May 2020