Hi, I'm David Rosales

Computer Science graduate from the University of Wisconsin-Madison

Building efficient and scalable software solutions.

About Me

Photo of David Rosales

I'm a Computer Science graduate from the University of Wisconsin-Madison with a strong interest in software engineering, machine learning, and cybersecurity. I enjoy working on problems where I can combine logical thinking with real-world impact.

During my time at UW-Madison, I've worked on projects involving data structures, algorithms, artificial intelligence, and natural language processing. I'm especially interested in how software systems are designed, optimized, and scaled in practice.

Outside of academics, I've gained experience working in both technical and customer-facing roles, which has helped me develop strong communication skills and a collaborative mindset.

Projects

Toxic Comment Classification

June 2025 - August 2025

Python · NLP · Machine Learning · Naive Bayes · TF-IDF

  • Built a multi-label toxicity classification system using the Jigsaw dataset (~159K comments) to detect six categories of online abuse.
  • Designed an NLP preprocessing pipeline with tokenization, lemmatization, and TF-IDF feature extraction.
  • Trained independent Multinomial Naive Bayes classifiers to balance interpretability, scalability, and inference speed.
  • Achieved an average accuracy of 97.2%, with threat detection reaching 99.4% accuracy.
  • Evaluated tradeoffs between probabilistic and neural approaches, prioritizing models suitable for real-time moderation systems.
  • Analyzed model limitations in detecting implicit toxicity and sarcasm, identifying areas for future improvement.

Experience

Configuration Technician I · CDW

July 2022 - July 2023
  • Deployed software and system updates to laptops, desktops, and mobile devices using secure company VPNs and internal configuration tools.
  • Ensured all devices met customer and quality standards prior to shipment in a high-volume production environment.
  • Worked with enterprise hardware including Kronos time clocks (Chipotle), Epson receipt printers (Ulta Beauty), and VR headsets (Medtronic).
  • Trained new hires on configuration workflows and collaborated with team members to meet tight deadlines.

Contact

Feel free to reach out, I'm always open to connecting or collaborating.

Contact Me