|

Peak productivity hours. Let's talk.

Building tools that optimize decisions.

Full-Stack Developer • Machine Learner • Houston, TX

What I Do

Options, not decisions.

f(x)

Optimization Algorithms

Linear programming and constraint solving to find mathematically optimal solutions to real problems.

┌|∵|┘

Machine Learner

From tree models to fine tuning LLMs. Understanding why models work is my thing, not just that they work.

</>

Full Stack Dev

Next.js, FastAPI, PostgreSQL. Learning to build complete systems from database to deployment.

||:

Data Pipelines

Web scraping, API integration, real-time data processing. Getting the data that matters.

LIVE STATS

Always building.

...

WPM

MonkeyType 60s

...

Day Streak

GitHub

...

Contributions

This Year

...

Accuracy

Best Test

♪ spotify: offline

THE PROGRESSION

Each project builds on the last.

MAY 2025

Spotify Clone->

Mastering React and Tailwind CSS fundamentals

ReactTailwind CSSREST APIs

NOV 2025

FullCircle->

Applying ML to improve music recommendations

Next.jsXGBoostRandomForest

NOW

SSMLA->

Fine tuning models with custom teaching data

Phi-3LoRAHugging Face

NEXT

Personal AI System

Orchestrating multiple models for personal workflows

Multi-modelAutomationIntegration

FLAGSHIP PROJECT

Cuenta. Grocery Intelligence.

Optimizes grocery shopping for protein efficiency within budget constraints.

LP

Linear Programming

PuLP optimization engine

100+

Products Tracked

Real Houston store data

99%

Budget Efficiency

Optimal utilization

80+

Houston Stores

Kroger, HEB, Target, Aldi

* * * * * * * *

CUENTA

GROCERY INTELLIGENCE

KROGER GREEK YOGURT$0.89
17g protein19.1 g/$
CHUNK LIGHT TUNA$0.99
29g protein29.3 g/$

THE DATA IS YOURS

the decision should be too

THE JOURNEY

From Research to Reality.

It started with a research paper. It transitioned into a passion. After years of studying different paths, 2025 showed me no matter how much I try, the only person who's in control of my fate is myself. Now all that's left to do is create.

FullCircle was the testing ground. A Spotify recommendation engine. Originally I considered RandomForest, but XGBoost did a couple things better— the heavy lifting. The model worked. More importantly, I understood why.

SSMLA is the next logical step. Not just using libraries, but hands-on learning. Phi-3 first, lighter and forgiving with a large support system. A developmental "shot in the dark", and at the moment still in the building phase.

The endgame? My own multi-agent system (MAS). SSMLA is the stepping stone, just as every other project has been to date.

[1]

Tree Models

[2]

XGBoost

[3]

Single Model

[4]

Multi Agent

FOR MY PEOPLE

Skills mean nothing without impact.

Building for my communities at UH. Real users, real problems, real accountability.

HISPANIC BUSINESS STUDENT ASSOCIATION

ALPFA Points->

ACTIVE

Gamified member engagement system with racing-themed familia competition. Because tracking points in a spreadsheet wasn't cutting it.

Familia leaderboards with real time rankings
Duo 2FA using UH's existing system
ColorStack-inspired activity streaks
DjangoTailwindPostgreSQLDuo API

"La familia siempre gana"

IEEE @ UNIVERSITY OF HOUSTON

Inventory System

PLANNED

ML-powered tracking system for 500+ electronic components. Predicting hardware needs before they become shortages.

Demand prediction based on event calendar
$10K+ annual budget optimization
Location mapping for 100+ engineers
FastAPIReactPostgreSQLscikit-learn

Extending SSMLA's prediction capabilities to real inventory data

Both projects will serve 1000+ students at University of Houston

LET'S TALK

Ready to connect?

Available for technical discussions, project walkthroughs, or just to chat.

Available this month

I type at 159 WPM with 100% accuracy|

* live from MonkeyType