Home About Projects Contact

Resume

Home About Projects Contact

Resume

Clutch Studio

Loading your experience

We are preparing the page. This only takes a moment.

Resume & Research

Resume

Socials

Instagram

Facebook

Let's Connect

Email:

clutchdev.apps@gmail.com

Phone:

949-910-7879

Quick Links

GitHub Terms of Service Privacy Policy Admin Login

Clutch Studio

Loading your experience

We are preparing the page. This only takes a moment.

Home About Projects Contact

Resume

Home About Projects Contact

Resume

San Diego, CAUCSD · B.S. CogSci (ML/AI) + CS Minor

Hi, I'm Ramon

I'm glad you're here! Look around to see my Full Stack and Machine Learning projects 😊.

I build products end-to-end: front-end polish, scalable backends, and applied ML. Focused on performant, human-centered experiences.

Projects Resume ML Research

San Diego, CAUCSD · B.S. CogSci (ML/AI) + CS Minor

Hi, I'm Ramon

I'm glad you're here! Look around to see my Full Stack and Machine Learning projects 😊.

Projects Resume

Scroll down to view recent project

Resume

Recent Projects

Clutch Studio

Loading your experience

We are preparing the page. This only takes a moment.

Resume & Research

Resume

Socials

Instagram

Facebook

Let's Connect

Email:

clutchdev.apps@gmail.com

Phone:

949-910-7879

Quick Links

GitHub Terms of Service Privacy Policy Admin Login

San Diego, CAUCSD · B.S. CogSci (ML/AI) + CS Minor

Hi, I'm Ramon

I'm glad you're here! Look around to see my Full Stack and Machine Learning projects 😊.

I build products end-to-end: front-end polish, scalable backends, and applied ML. Focused on performant, human-centered experiences.

Projects Resume ML Research

San Diego, CAUCSD · B.S. CogSci (ML/AI) + CS Minor

Hi, I'm Ramon

I'm glad you're here! Look around to see my Full Stack and Machine Learning projects 😊.

Projects Resume

Scroll down to view recent project

Resume

Recent Projects

Comparing the performance of different Classification and Regression Algorithms

AI•Machine Learning•Python

I built a comprehensive machine learning pipeline for tabular data and for both regression and classification tasks that compared the performance of Gradient XGBoosting, Random Forest, Neural Networks, and SVMs (though SVMs are excluded from the paper due to time constraints in hyperparameter tuning especially for large datasets for it to be competitive, but you can try it yourself). We compared three different splits 20/80, 50/50, and 80/20 across three independent trials with different seeds using 10-Fold stratified CV for classification and 10-Fold CV for regression for each model and each dataset. The results were very close, but the best model was Gradient Boosting as it had the best F1 score for two out of the three classification datasets and the best RMSE score for one out of the two regression datasets. This research paper followed the NeurIPS 2023 format and was inspired by Richard Caruana and Alexandru Niculescu-Mizil in their paper An Empirical Comparison of Supervised Learning Algorithms. The paper is not yet submitted for publication, but with some revision like adding more datasets and fully integrating the SVM into the paper, it could be submitted for publication.

View Code View Project

Built a feature rich hackathon and ideathon platform for student orgs and companies.

Startups•AWS S3•Web Development

I built a full-featured hackathon and ideathon platform for student organizations and companies to host, manage, and scale their competitions. The platform is data-driven and designed around an intuitive end-to-end workflow, including organization management, event creation, team formation, built-in matchmaking, submissions, judging, and live leaderboards. It supports multiple event types, role-based access control, and real-time state changes across the entire judging and submission pipeline. The platform is built with Next.js 16, React 19, TypeScript, and Tailwind CSS, with AWS S3 handling media storage. I used Prisma ORM for database modeling and authentication, and implemented custom server actions for form handling, permissions, and transactional workflows.

Visit Website View Project

Built a comprehensive e-commerce platform to expand STEM education and technology access to underprivileged communities

Web Development•Next.js 15•React

I developed a full-stack e-commerce platform for my nonprofit foundation dedicated to expanding STEM education and technology access to underprivileged communities. The platform features a complete shop system with print-on-demand merchandise through the Printful API, a donation system with both one-time and recurring contributions via Stripe, and automated email notifications using the Resend API. I implemented NextAuth.js v5 for secure authentication, built a comprehensive order management system with webhook handlers for both Stripe and Printful, and created a partnerships program to facilitate organization collaborations. The platform includes guest checkout support, persistent shopping carts, real-time inventory management, and a social sharing system to amplify the mission. With 50% of all profits directed to supporting digital equity initiatives, the website serves as both a revenue generator and community hub for creating creators rather than renters in the digital space.

Visit Website View Project

Trained a custom YOLOv8 model to identify 9 specific desk objects with 80% training accuracy

AI•Object Detection•YoloV8

I created a custom YOLOv8 object detection model trained on 351 hand-labeled images across 9 desk object classes. The project involved recording 5 training videos from multiple viewpoints (birds-eye, front, horizontal, left, right), extracting frames every 2 seconds, and manually annotating each image. I implemented a complete pipeline including JSON-to-YOLO format conversion, dataset splitting (80% train, 20% validation), and model training for 84 epochs. The final model achieved a fitness score of 0.7944 and successfully detected all target objects in real-world video scenarios after confidence threshold optimization.

View Code View Project

Comparing the performance of different Classification and Regression Algorithms

AI•Machine Learning•Python

I built a comprehensive machine learning pipeline for tabular data and for both regression and classification tasks that compared the performance of Gradient XGBoosting, Random Forest, Neural Networks, and SVMs (though SVMs are excluded from the paper due to time constraints in hyperparameter tuning especially for large datasets for it to be competitive, but you can try it yourself). We compared three different splits 20/80, 50/50, and 80/20 across three independent trials with different seeds using 10-Fold stratified CV for classification and 10-Fold CV for regression for each model and each dataset. The results were very close, but the best model was Gradient Boosting as it had the best F1 score for two out of the three classification datasets and the best RMSE score for one out of the two regression datasets. This research paper followed the NeurIPS 2023 format and was inspired by Richard Caruana and Alexandru Niculescu-Mizil in their paper An Empirical Comparison of Supervised Learning Algorithms. The paper is not yet submitted for publication, but with some revision like adding more datasets and fully integrating the SVM into the paper, it could be submitted for publication.

View Code View Project

Built a feature rich hackathon and ideathon platform for student orgs and companies.

Startups•AWS S3•Web Development

I built a full-featured hackathon and ideathon platform for student organizations and companies to host, manage, and scale their competitions. The platform is data-driven and designed around an intuitive end-to-end workflow, including organization management, event creation, team formation, built-in matchmaking, submissions, judging, and live leaderboards. It supports multiple event types, role-based access control, and real-time state changes across the entire judging and submission pipeline. The platform is built with Next.js 16, React 19, TypeScript, and Tailwind CSS, with AWS S3 handling media storage. I used Prisma ORM for database modeling and authentication, and implemented custom server actions for form handling, permissions, and transactional workflows.

Visit Website View Project

Built a comprehensive e-commerce platform to expand STEM education and technology access to underprivileged communities

Web Development•Next.js 15•React

I developed a full-stack e-commerce platform for my nonprofit foundation dedicated to expanding STEM education and technology access to underprivileged communities. The platform features a complete shop system with print-on-demand merchandise through the Printful API, a donation system with both one-time and recurring contributions via Stripe, and automated email notifications using the Resend API. I implemented NextAuth.js v5 for secure authentication, built a comprehensive order management system with webhook handlers for both Stripe and Printful, and created a partnerships program to facilitate organization collaborations. The platform includes guest checkout support, persistent shopping carts, real-time inventory management, and a social sharing system to amplify the mission. With 50% of all profits directed to supporting digital equity initiatives, the website serves as both a revenue generator and community hub for creating creators rather than renters in the digital space.

Visit Website View Project

Trained a custom YOLOv8 model to identify 9 specific desk objects with 80% training accuracy

AI•Object Detection•YoloV8

I created a custom YOLOv8 object detection model trained on 351 hand-labeled images across 9 desk object classes. The project involved recording 5 training videos from multiple viewpoints (birds-eye, front, horizontal, left, right), extracting frames every 2 seconds, and manually annotating each image. I implemented a complete pipeline including JSON-to-YOLO format conversion, dataset splitting (80% train, 20% validation), and model training for 84 epochs. The final model achieved a fitness score of 0.7944 and successfully detected all target objects in real-world video scenarios after confidence threshold optimization.

View Code View Project