Hi!
Nice to meet you.
I'm Miguel.

I'm an aspiring software engineer based in the Bay Area. I love bringing ideas to life through code. Whether it's coding up a new sleek interface or writing new robust APIs, I'll always be looking out for new challenges to improve myself.

Check out my work!

A picture of myself

My Projects

Check out some of the things I built!

An app focused on discovering third spaces

Coffee Spot

An app focused on discovering third spaces

This was one of my favorite projects. Here I got the opportunity to become team lead for our capstone project. blah blah blah

Technologies used:
Next.jsMongoDBNode.jsTypescriptTailwindPrisma
Made using the websocket API

Chat App Application

Made using the websocket API

I built a real-time messaging application as a project for learning more about the websocket API. There's features such as a friend management system, online status tracking, group messaging, and direct messaging

Technologies used:
React.jsPythonFastAPITypescriptAWS S3
A personal online teaching assistant

TA.ai (SF Hacks 2025)

A personal online teaching assistant

For SF Hacks 2025, three of my friends and I built an online platform to help teachers generate educational worksheets with adjustable difficulty levels. Teachers and professors can submit learning material to the website and recieve customizable generated worksheets for students

Technologies used:
Next.jsSupabaseTypescriptNextAuthGoogle GeminiTailwind CSS
For a more streamlined LaTeX experience

Handwritten Equation to LaTeX Converter

For a more streamlined LaTeX experience

Since we hated having to manually type matrices into LaTeX, we trained an OCR to convert the math equations in your notebooks to LaTeX for online homework submissions. Using YOLOv8,a dataset of 10,000 handwritten calculus equations, and some preprocessing techniques, we were able to achieve a 94% validation accuracy.

Technologies used:
PythonYOLOv8PyTorch
Helping new baristas learn new drinks quicker

Barista Trainer

Helping new baristas learn new drinks quicker

During my time at Starbucks, I created a web app that helped new baristas memorize our 150 drinks. It utilized Anki's SRS algorithm for quiz mode and even timed mode which times how fast you can get a drink ready in time.

Technologies used:
PythonYOLOv8PyTorch

Experience

Here are some of the places I worked at.

Xoloosh - Software Engineer Intern

June 2025 - Present

  • Contributed to the migration of a legacy .NET financial application to a modern Django and Vue.js architecture, supporting the redesign of expense tracking systems used by multiple community college districts to manage budgets, contracts, and vendor contacts
  • Modernized an outdated Javascript Gantt chart library by refactoring legacy code to modern ES6 classes, then redesigned and rebuilt the component from scratch using web components for improved portability and framework compatibility, including a Vue.js wrapper for seamless integration.

Contact

Interested? Let's connect!

Like what you see? Let's connect