Software Engineer & AI/ML Engineer

Alfonso
De La Cruz

MS Computer Science graduate building intelligent software systems. Passionate about AI/ML engineering and data pipelines.

Python Machine Learning LLM Fine-Tuning C++ JavaScript React SQL Firebase
View My Work Get In Touch
Scroll to explore
01

About Me

I'm a Computer Science Master's graduate from California State University San Marcos, where I also completed my BS with a 3.5 GPA and made the Dean's List four times.

My thesis focused on network traffic anomaly detection using machine learning — comparing Random Forest, Logistic Regression, and SVM with advanced sampling techniques on real-world botnet datasets.

I'm passionate about the intersection of software engineering and AI — building systems that don't just work, but learn. I approach problems with curiosity, ship fast, and iterate on what I learn.

Skills & Technologies
Languages
Python C++ JavaScript SQL Java / Kotlin
AI / ML
PyTorch Hugging Face Transformers scikit-learn LLM Fine-Tuning SMOTE / Resampling Feature Selection
Frameworks
React JUnit Firebase Hugging Face
Tools
Git VS Code PyCharm Eclipse Google Colab
3.5
GPA — BS Computer Science, CSUSM
4×
Dean's List recipient
MS
Computer Science — CSUSM 2026
80%
Accuracy on unseen commands — LLM fine-tuning proof of concept
02

Projects

MS Thesis — Capstone Project

Network Traffic Anomaly Detection
Using Machine Learning

Detecting botnet activity in real network traffic using the CTU-13 dataset — a benchmark dataset of real malware captures from Czech Technical University. Compared three classifiers under severe class imbalance conditions, defended May 2026.

3
ML models compared
2
Sampling strategies (SMOTE + undersampling)
13
CTU-13 network scenarios
4
Feature selection methods
Problem

Network traffic datasets are severely imbalanced — botnet traffic is rare but dangerous. Standard classifiers fail to detect minority class anomalies without targeted resampling strategies.

Approach

Applied SMOTE (synthetic oversampling) and random undersampling across Random Forest, Logistic Regression, and Linear SVC. Used Chi-Square, ANOVA, RFE, and embedded feature selection to reduce dimensionality.

Key Finding

Random Forest with SMOTE consistently outperformed other combinations, achieving the best balance between precision and recall on botnet detection. Results benchmarked against Sharma et al. (2024) who used XGBoost on the same dataset.

Python scikit-learn SMOTE Random Forest Linear SVC Logistic Regression ROC Curves CTU-13 Dataset Feature Selection Google Colab
View Full Thesis →
AI / ML — Personal
LLM Fine-Tuning for Home Automation
Fine-tuned GPT-2 on home automation commands to classify natural language intent into structured device actions. Achieved 80% accuracy on unseen phrases. Measured and optimized inference latency by 63% through greedy decoding and token reduction.
PyTorch Hugging Face GPT-2 Latency Optimization
Web — Personal
Higher or Lower
Fan-built interactive web game. Players guess tournament placements across seasons. Features split-screen card layout, streak counter, and 3-life system.
HTML/CSS/JS Netlify Game Design
03

Experience

Mar 2025 — Jul 2025
Information Systems & Technology Intern
Escondido Union School District
Maintained and updated EUSD's public-facing website, ensuring content accuracy and timely publication of district announcements. Audited website content for ADA compliance, resolving accessibility violations across multiple pages. Implemented semantic HTML structures to ensure compatibility with screenreaders and ADA-accessible devices.
HTML/CSS ADA Compliance Web Development Accessibility
Aug 2024 — May 2026
Instructional Student Assistant
California State University San Marcos
Tutored students enrolled in Operating Systems, Data Structures, and Algorithm Design — reinforcing core CS concepts through one-on-one and group sessions. Assisted professor with grading assignments and exams for a class of 30+ students, providing detailed written feedback. Tracked and maintained grade records for all enrolled students throughout the semester.
Operating Systems Data Structures Algorithms Teaching Mentoring
2025 — 2026
President — AI Club
California State University San Marcos
Led development of Project Mindcraft — deploying AI agents in Minecraft using Python and open-source LLM frameworks. Delegated tasks across a student team and coordinated development sprints. Organized club meetings, technical workshops, and speaker events to grow student engagement in AI and machine learning.
Leadership Python LLM Frameworks Public Speaking Team Management
04

Education

2024 — 2026
MS Computer Science
California State University San Marcos
Thesis: "Network Traffic Anomaly Detection Using Machine Learning." Compared Random Forest, Logistic Regression, and SVM on CTU-13 botnet dataset with SMOTE and undersampling techniques. Random Forest achieved F1 = 0.9965 — outperforming state-of-the-art XGBoost baseline on the same dataset. Defended May 2026. View thesis →
3.7 GPA Machine Learning Python scikit-learn
2020 — 2023
BS Computer Science
California State University San Marcos
Graduated with 3.5 GPA. Dean's List four times. Coursework in AI (search algorithms, Prolog expert systems, predicate calculus), database systems, design patterns, and software engineering. Built multiple full-stack and mobile applications throughout the program.
3.5 GPA Dean's List ×4 C++ Java AI / Prolog
05

Contact

Let's build something together.

I'm currently open to full-time software engineering and AI/ML engineering roles. Based in San Diego, CA — open to relocation for the right opportunity.

✉
Email
alfonsojd11@gmail.com
in
LinkedIn
linkedin.com/in/alfonsojd