Hello!

Hello! My name is Keegan, and I am a full stack web developer and musician. I recieved a Doctorate of Musical Arts from the University of Southern with a major in Studio/Jazz Guitar. After working as a musician and adjunct music instructor since 2013, I decided to switch my focus to software developement.

I completed a full stack web development bootcamp through UCLA Extension in 2022, and I have been programming daily ever since. I competed in several hackathons including: Chainlink (2022, 2023) as a smart contract developer, DearJouniorDev (2023) as a full stack developer, and EdgeDB a backend developer. Over the course of 2023 and 2024, I have worked on two major projects.

The first project was called pocketPR, which is a web application that allowed content creators to connect their YouTube account to the app, and use an assistant to help to mitigate conflict within their audience. We did this by supplying the AI assistant with video transcripts, and comment sections from You Tube. My friend Daniel Johnson and I built this app over the course of 10 months with NextJS, Supabase, Clerk, Openai API, You Tube Data API, and deployed to Vercel.

My current project is called chatty-chapters.click which is a web application that I am using in my music classes at College of the Desert and Chaffey College. Since ChatGPT came into the mainstream in 2023, I, like make college educators, struggled with how to deal with the increased usaged of LLMs on class assignments. Should we outlaw all usage of them? Try to police the use of them? How can you even grade assignments fairly when one student's answer might be very impressive, but obviously written by a machine? Then, you have another student who maybe didn't try that hard on the assignment, but at least they wrote it themselves.

chatty-chapters.click has been my attempt to solve this problem. I pre-populate the app with the course textbooks, and then I have the students chat with an AI about the chapter. The AI has access to the course textbook via embeddings, which makes it possible to do semantic search and provide the AI with the proper context to answer the student's questions. Next, the student takes a quiz that has been generated by the AI, based on the chapter, and the conversation. The AI will also grade the quiz, and provide the student with feedback. As an instructor, I also have a dashboard that allows me to view all student-AI interactions. I built this app with NextJS and Go and deployed it to AWS using a variety of services including: S3, EC2, Cognito, and Amplify.

I continue to work as an adjuct instructor of music at College of Desert and Chaffey College, and I also teach Intro to Web Design at Cypress College. I am interested in working on interesting projects that can have a positive impact on peoples lives. Bonus if the project involves music and/or education!

In terms of technologies, I have built projects with JavaScript/TypeScript, Go, and Python. I have a preference for working in Go, as I enjoy the simplicity of the language, error handling, etc. This site's backend was made with a Go backend, and good old fashion HTML/CSS/JS for the front end. For most sites that I have build however, I tend to use NextJS as a the front end framework.