FUTO is a software company devoted to putting control of our computers back in the hands of the people. I work on the FUBS project, a minimal software development system that is easy to understand, modify, and extend, with an emphasis on embedded devices. My main focus is on OS- and runtime-level components, such as memory allocation and device drivers.

I build and maintain backend services for Notifi, a privacy-respecting messaging platform for Web3 and DeFi services. My main focus is on the security and stability of the Notifi platform, including automated controls to maintain high availability to both Notifi's systems and our messaging and blockchain API partners.

I contribute to Clusterfun, a video game website featuring HTML5/React games controlled with mobile devices. My main contributions include the communication system (utilizing Node.js, Express, and WebSockets) that keeps the player devices and central presenter in sync. I also help build and playtest the games themselves.

I built and maintained backend services at Amazon Luna, utilizing agile methodology, Java, Node.js, and AWS distributed systems (with both EC2- and Lambda- based services, using IAM, SQS, SNS, S3, and DynamoDB to store data and transmit it between systems) to deliver highly available and scalable systems to enhance the customer's social and game recommendation experience. I also built and maintained backend systems for Amazon Prime Video, using the same set of skills to deliver highly scalable and reliable systems capable of serving and managing metadata for millions of unique video titles.

As an item of particular interest, I was asked to mediate the addition of a quality control check in Prime Video's internal content submission systems, in order to correct errors on the front page. The team that owned the codebase that was to implement the check disagreed with it, so in the interest of objectivity, in addition to implementing the change, I also studied the deficiency in detail to figure out how much effect the check would have - while I determined that implementing it would be worthwhile, I identified that over 50% of the observed errors were due to employee error later down the submission pipeline, and saw a significant improvement in the error rate after those issues were corrected.

I built custom production management software for Pyxis Manufacturing, a textile company that primarily produces custom-monogrammed products. The software used React, Node.JS, MongoDB, and Docker to manage inventory and order progress as well as collect data on factory efficiency. We also used AWS to manage the software's connection to remote workers, such as designers.