Portrait of Kai Morales

Kai Morales

Undergraduate · University of Pittsburgh · B.S. Data Science

About

I’m a college student focused on data science and applied machine learning. I like shipping open-source tools—especially in sports analytics or anything competitive.

Projects

CFB Rank — College Football Ranking Algorithm

Aug 2025 – Present · cfbrank.org

Developed a Python-based college football ranking algorithm that replaces subjective AP/Coaches poll voting with a quantitative framework using the CFBD API, pandas, and data-driven weighting. The model achieved results that outperform legacy polls in predictive validity — accurately flagging over-ranked teams like Texas and Penn State before their Week 6 losses to unranked opponents.

ClearFit — AI Resume Analyzer

Mar 2025 – Aug 2025 · clearfit.tech

Full‑stack app that delivers structured, no‑fluff resume feedback aligned to a job post. Implemented auth, payments, and secure uploads; kept a private instance to control API costs.

MNIST Digit Classifier (PyTorch)

PyTorch · simple CNN

Built a modular neural network in PyTorch to classify handwritten digits (0–9), achieving ~98% accuracy on the MNIST dataset. Implemented clean data loaders, training loops, dropout regularization, and model persistence for reproducibility.

Experience

AI Research Intern — FACET Lab

University of Pittsburgh · Sep 2025 – Present

  • Contribute to BRIDGE (NSF) building inclusive AI/AR tools for deaf learners in collaborative STEM.
  • Backend services for Unity (C#): data models and real‑time pipelines for sign detection and overlays.
  • Prototype real‑time ASL recognition (CV + basic NLP) with on‑device feedback in Unity.

Data Analytics Intern — The Pitt News

Sep 2025 – Present

  • Launched a GA4‑backed analytics dashboard used weekly by 25+ editors/marketers.
  • Added plain‑English explainers and next‑steps, cutting time‑to‑insight from ~20m to ~6–8m.
  • Recommendations on send times/headlines contributed to ~22% CTR lift on optimized campaigns.

ML Robotics Operator — Stealth Startup ($4.5 Billion)

Dec 2024 – Present

  • Operate & troubleshoot AI robotics systems to accelerate model‑training pipelines.
  • Developed a more efficient physical task‑reset method (~20% faster training cycle).
  • Document anomalies and collaborate with engineers to improve behavior and accuracy.

Education

University of Pittsburgh — B.S. Data Science (Minors: Computer Science & Applied Statistics), expected May 2028

Relevant coursework: Intermediate Programming; Data Structures & Algorithms; Data Science; Assembly & Computer Organization; Calculus; Matrices & Linear Algebra; Applied Statistical Methods; Discrete Mathematics.

Skills

Python, Java, R, PyTorch, pandas, scikit‑learn, SQL, Linux, HTML/CSS, Firebase, GA4, Tableau, Markdown.

Contact

kaimor06@gmail.com · 202‑805‑2836 · LinkedIn ·