Kylie Cancilla

Interests

Deep Learning Architectures , AI Innovations , Transfer Learning in Medical Applications , Predictive Modeling for Disease prevention

Skills

Languages and Frameworks: Python (PyTorch, Pandas, NumPy, Scikit-learn, Matplotlib), Java, R, HTML, SQL

Education

University of San Francisco

B.S., Data Science (expected May 2026)

2022 - Present

Work

Data Science Fellow, Lawrence Livermore National Laboratory

Contributed to Computational Visual Perception Research

  • Developed and trained PyTorch-based neural networks to predict image quality of video frames, enhancing video object detection performance.
  • Collaborated in a data science challenge with fellow interns, designing and training models to predict amodal masks for objects in video sequences.

2025 - Present

Data Science Intern, Women in Data Science Worldwide

www.widsworldwide.org/

Assisted in organizing the 2025 Women in Data Science (WiDS) annual datathon.

  • Developed a Jupyter Notebook and led workshops for 100+ participants to predict ADHD in adolescent girls using socio-demographic, parenting and connectome data. Guided data exploration, cleaning, and multi-output machine learning model development for ADHD diagnosis and patient sex prediction.

2024 - 2025

Awards

Presidential Scholarship, University of San Francisco

2022-present

Dean's Honor Roll - 4.0 GPA, University of San Francisco

2022-present