Mon Fernandez

I’m a Computer Science student focused on building practical software in real codebases and improving generative AI output quality with disciplined evaluation. I enjoy turning requirements into working features—then validating results with clear, repeatable checks.

In an Open Energy Dashboard CTI-CodeDay SWE Micro-Internship, I contributed to an open-source system that converts and displays collected energy readings. I configured a unit-conversion test case in the JavaScript codebase and learned TypeScript to meet contribution requirements.

In my UCSB CS156 project, I worked inside an existing Java and JavaScript codebase using React to deliver a web application showing UCSB dining menu options. I participated in debugging and implemented functionality alongside mutation tests, using GitHub version control as part of the course workflow.

For my independent AI project, I developed and iterated prompts by comparing outputs across repeated tests, then ranked results using a consistent rubric and written justifications. I also trained LoRA adapters with Kohya-SS / sd-scripts—curating datasets and tuning settings in GPU workflows (e.g., Google Colab)—so the outputs become more consistent, clearer, and less artifact-prone.