Hey!👋 My name is

Cyrus Navasca

I am a Data Scientist based in the Bay Area. Through both industry and research experience, I have developed Agentic AI and Machine Learning solutions.

I LOVE to talk, so feel free to reach out on my contact page and grab a copy of my resume below on your way out!

Download Resume
Cyrus Navasca
Wild Pokémon bothering me on my walk 😞
Teaching Python at DS Collab
Teaching some poor souls about Python
Milk challenge
"I love taking on new challenges!" (I took on a gallon of milk once... and lost)

Projects

Acuity.ai ED Risk Assessment dashboard

Acuity.ai

  • Developed an AI clinical assistant which helps assess a patient's risk of readmission to the Emergency Department.
  • Leveraged regex-based NLP to parse medical documents and ensemble modeling generate accurate predictions.
  • Architected a RAG pipeline to surface relevant medical knowledge for chatbot responses
  • Deployed using FastAPI to manage patient data, predictions and chatbot states
GitHub
Diabetic retinopathy detection model predictions

Detecting Diabetic Retinopathy w/ Computer Vision

  • Architected an end-to-end deep learning pipeline using PyTorch to detect diabetic retinopathy from retinal images.
  • Streamlined data management by utilizing AWS to store over 3,000 retinal scans.
GitHub
Data 4 Dummies workshop

Data 4 Dummies

  • Formulated a beginner-friendly data science workshop series covering topics from basic Python programming to machine learning.
  • Led over 350 members through coding worksheets using sample datasets and presentation slides to facilitate learning.
Visit

Skills

Languages/Tools

Python
Python
AWS
AWS
Git
Git
GitHub
GitHub
SQL
SQL
Tableau
Tableau
Excel
Excel
R
R

Data Science/ML

PyTorch
PyTorch
LangChain
LangChain
scikit-learn
scikit-learn
ZenML
ZenML
MLFlow
MLFlow
PostgreSQL
PostgreSQL
MongoDB
MongoDB

Experiences

Data Science Intern

June 2025 – Present

Elas

  • Engineered an AI forecasting agent to turn natural language queries to actionable insights leveraging LangChain and ensemble modeling
  • Integrated Vanna AI with PostgreSQL and AWS to enable clients to convert prompts into optimized SQL queries
  • Reduced LLM token costs by 65% and response latency by 90% for an AI data visualization agent by optimizing its data pipeline to extract essential metadata
Visit

Data Science Research Assistant

Feb 2025 – Nov 2025

META Lab

  • Developed a natural language processing workflow leveraging Sentence Transformers to generate embeddings, boosting predictive model accuracy by 58%.
  • Constructed a segmentation model by applying PCA and K-means clustering to identify "low-use" and "heavy-use" cannabis users.
  • Spearheaded statistical analysis to uncover relationships among curiosity, creativity, and intervention types using multivariate regression, mediation models, ANCOVA and A/B testing
Visit

Data Analyst Intern

Oct 2024 – May 2025

Daily Nexus

  • Implemented an end-to-end data workflow leveraging Selenium to scrape 3,000+ websites
  • Published an interactive article analyzing trends between Rate My Professors ratings and UCSB grade distributions
Visit Publication

* Runner-up for CCMA Best Interactive Graphic

Education

Campus

University of California, Santa Barbara

Santa Barbara, CA

B.S. Statistics & Data Science, Minor in Mathematics

Relevant coursework: Advanced Linear Algebra, Statistical Machine Learning, Time Series Analysis, Applied Stochastic Processes, Bayesian Data Analysis

Contact

I can be reached at cyrus.navasca54@gmail.com or through the form below.