Bulldozer Price Regression
End-to-end ML pipeline on 401K+ Blue Book auction records. Cleaned data, engineered features, and compared Linear/KNN/Decision Tree/Random Forest; best model achieved R² = 0.91 with GridSearchCV tuning.
Carlos Felipe
CS Engineering Student • Software Projects • Open to Opportunities
Skills
Languages
C++
Python
JavaScript
SQL
HTML/CSS
Spanish (native)
English (fluent)
Frameworks & Tools
Django
React
PostgreSQL
Tailwind CSS
Vite
Git & GitHub
Data & ML
Pandas
NumPy
scikit-learn
Matplotlib
Seaborn
Excel
Coursework
Data Structures
Software Engineering
Database Systems
Data Science
Intro to Machine Learning
Operating Systems
Natural Language Processing
Projects
EventSync
Full-stack event platform built with Django, PostgreSQL, and Tailwind. Users can create, browse, and purchase event access; includes secure auth, scalable DB (≈1,000 events/users), real-time chat, and Stripe payments.
Customer Segmentation in Retail
Predicted purchase likelihood on the Online Retail dataset. Engineered recency, transaction count, and revenue features; evaluated Logistic Regression vs. Random Forest (balanced data ~50% accuracy). Delivered targeted promo insights.
Fast-Track
Route-optimization web app with Dijkstra’s algorithm and a custom MinHeap from scratch; visualizes and compares Dijkstra vs. A* on a weighted U.S. cities graph.