Dana Zarezankova Email & Phone Number

• Implemented batch processing for SpaCy NLP part of speech tagging and PySpark to reduce compute time by 60%, patent in progress.
• Fine-tuned PyTorch model to enhance precision and recall for identifying high insurance cost patients.
• Gen AI for personalized physiotherapy educational articles with few-shot prompting via LangChain.

• Worked on medical record matching systems using random forests, XGBoost, and logistic regression models.
• Authored a Python library to facilitate the export of scikit-learn models to JSON format for cross-platform deployments.
• Authored Python libraries for phonetic processing of medical record data for enhanced matching accuracy.

• Worked on a regression model designed to predict egg retrieval outcomes.
• Conducted competitive analysis using Monte Carlo simulations to estimate the error margins in competitor models based on their published research.
• Implemented an automated data collection pipeline from the SART online birth rate predictor tool to supplement our existing patient success dataset using Selenium for web scraping.

• Lead the first iteration of an intelligent in-vitro fertilization clinic finder for patients.
• Scraped and cleaned clinic success and patient statistics from the CDC for a clinic ranking system.
• Automated web scraping to collect patient review information on doctors and clinics for clinic profile pages.

Toronto, Ontario, Canada

• Built and shipped customer loyalty algorithms for Shopify apps in production, serving over 100 stores over 8 months.
• Developed an API that emails targeted marketing campaigns to unique customer segments using AWS Lambdas and Jenkins.

Toronto, Ontario, Canada

- Tested medical grade software in an agile software development team.- Wrote end to end automated testing scripts for all user flows.- Created medium fidelity wireframes for app interface.- Documented verification protocol for medical grade software.