Hi, I'm Sonya
Research data scientist, currently pursuing an MS in Data Science at Boston University. I build predictive models, wrangle millions of clinical records, and genuinely believe data science can make healthcare better for the people who need it most.
What I'm exploring now
Updated May 2026
Deep in survival models. I've been studying the Anderson-Gill extension of the Cox model for recurrent events, along with joint frailty and shared frailty models. I used these to study recurrent exacerbations. The math is beautiful and the clinical applications are everywhere.
Part-time grad student life. Working toward my MS in Data Science at Boston University while continuing full-time research at Einstein. It's a lot, but I love it.
Reading everything from the VA research group. The work of Yan Xie, Taeyoung Choi, Ziyad Al-Aly, et al. on large-cohort post-COVID outcomes has been shaping how I address biases in observational studies.
About me
There is no single title that fully captures the scope of my work. I conduct research, perform data analysis, and build technical solutions, while remaining deeply engaged with the clinical aspects of healthcare. My role consistently bridges clinical research and data science, allowing me to translate complex healthcare data into insights that are both technically sound and clinically meaningful.
I am also a tutor, mentor, and lifelong learner. I have trained more than 90 researchers in data workflows and statistical methods and have developed educational resources to support reproducible and efficient research practices. My interest in continuous learning, particularly in emerging statistical and analytical approaches, has contributed to more than 20 peer-reviewed publications across a range of medical conditions and study designs.
I developed a passion for research since High school (The Manchester High School) where I was taught the fundamentals of scientific inquiry, this was further enhanced throughout undergraduate studies and my professional career. I am enjoying the learning process but have failed a lot more than I succeeded (Would I have it any other way? Honestly, yes but I am learning to fail forward).
Projects & Case Studies
Developed and fine-tuned large language models to assist with complex diagnostic cases, outperforming baseline algorithms. Explored prompt engineering, RAG pipelines, and evaluation frameworks for clinical accuracy.
Extracted, cleaned, and transformed millions of unstructured healthcare records into structured SQL databases. Created analysis-ready datasets and custom queries used across multiple concurrent studies.
Led a retrospective cohort study examining incidence rates and risk factors for pulmonary conditions up to 3 years post-COVID in an underserved urban population. First-author publication in Scientific Reports.
Publications · 23+ total peer-reviewed papers
What I'm reading
Research Group
I'm a big fan of thier work, especially their data analyses. Their large-cohort analyses on post-acute COVID outcomes have fundamentally shaped how I think about long-term disease burden and study design at scale. I especially admire how they control for biases. I learn something new every time I read their work.
Favorite Models
My favorite models to date. I used the Anderson-Gill extension of Cox regression to study recurrent exacerbations, and I've been drawn to joint frailty and shared frailty models for the way they account for unobserved heterogeneity across patients.
Interest Area
Following the rapid evolution of how language models are being evaluated and deployed in clinical settings. Thinking about where they genuinely help vs. where the hype outruns the evidence.
Interest Area
The populations I study are often the ones most underrepresented in research. In research, there should always be a balance between generalizability and the need to address disparities.
Skills & Tools
Before research
Before finding my way to research, I worked in customer service, tech support, and healthcare. Those roles taught me how to communicate clearly, troubleshoot under pressure, and build rapport with anyone. These experiences shaped how I approach collaboration and problem-solving today.
Community
Albert Einstein College of Medicine · 2022–Present
Mentoring researchers learning the craft, from data workflows and statistical methods to navigating the research process.
UStrive · 2020–Present
Helping high school students navigate college applications, essays, and the process of getting there.
American Red Cross · 2021–2022
Managed volunteer shift scheduling and coordination for community programs.
The City Tutors · 2020
Tutored Intro to Biology students weekly across two semesters.
Beyond the lab
Favorite classical pieces:
For a chill vibe:
I first played at 2 years old (taught by my dad). Currently on the hunt for a bass guitar.
Does not contribute to my publications but does provide moral support and the occasional keyboard interruption.