# RJ Ellis, PhD

> Sr. Director of Data Science @ RadiantGraph

Location: San Francisco Bay Area, United States
Profile: https://flows.cv/rj

I’m a data science technician and tactician in hot pursuit of ideas, methods, metrics, and tools that offer insights into what makes humans tick. My training in psychology, physiology, statistics, clinical neuroscience, and computer science has grounded my practice of translating complex ideas into straight-shooting insights in a way that’s persuasive, passionate, and personal. 

My approach is “problem-aimed, expert-informed, solution-focused” — helping stakeholders, clinicians, and end-users sift through big-deal data to discover answers to questions they didn’t know they could ask.

## Work Experience
### Senior Director, Data Science & Strategy @ RadiantGraph
Jan 2024 – Present | San Francisco Bay Area

### Director, Data Science & Strategy @ RadiantGraph
Jan 2023 – Jan 2024 | San Francisco Bay Area

### Manager, Data Science @ Teladoc Health
Jan 2020 – Jan 2023 | San Francisco Bay Area
→ Led the Connected Devices DS team (3 direct reports) and championed its mission: to better leverage raw and engineered signals from remote patient monitoring devices across the Teladoc ecosystem.

→ Deployed a daily pipeline that performs anomaly detection for 175K+ individuals with connected weight scales, ensuring high quality data for downstream DS applications (client reporting, clinical analytics, personalized models).

→ Launched a secure web portal that enables the DS team to design and deliver specific coaching interventions quickly and efficiently (i.e., with minimal engineering resources), accelerating delivery of cutting-edge human-powered coaching.

### Senior Data Scientist @ Livongo
Jan 2020 – Jan 2020 | San Francisco Bay Area
→ Data Science lead for internal projects related to continuous glucose monitoring, a major product initiative and key external partnership: opportunity sizing, power analysis, eligibility validation, clinical metric aggregation and ETL, longitudinal analyses.

→ Instrumented logic that enables validation of key self-reported health traits (e.g., diabetes type, insulin usage) data against external data sources (e.g., prescription fills, device readings), creating opportunities for proactive outreach to participants.

### Clinical Product Data Scientist @ Livongo
Jan 2018 – Jan 2020 | San Francisco Bay Area

### Fellow @ Insight Data Science
Jan 2018 – Jan 2018 | San Francisco Bay Area
→ Built a web app providing dynamically updated estimated-time-of-arrival predictions for 911-dispatched medical first responders in San Francisco (AWS + Dash). Details: http://robjellis.net/sf911.html

→ Wrangled open-source data sets (Google BigQuery) to develop a feature matrix comprising temporal, geospatial, and weather information associated with 460,000 distinct 911 medical calls between 2012 and 2017.

### Behavioral Researcher & Designer @ Omada Health
Jan 2016 – Jan 2017 | San Francisco Bay Area
→ Served as in-house subject matter expert on health behavior change, providing on-demand perspectives and analyses while working collaboratively with product design and user insights on product feature ideation, improvement, and implementation.

→ Co-authored a peer-reviewed journal paper detailing the three-year outcomes of Omada’s digital behavior change program for diabetes prevention; a first in the clinical literature. Details: https://drc.bmj.com/content/5/1/e000422

### Research Fellow in Computer Science @ National University of Singapore
Jan 2013 – Jan 2015 | Singapore
→ Proposed a method to quantify word choice novelty in pop song lyrics and released an open-access database of 275,000 lyrics, creating a best-in-class resource for the field of music information retrieval. Details: https://github.com/robjellis/lyricfind

→ Conceptualized and wireframed a mobile app, “SmartMOVE,” that provides hardware-free analytics of gait in patients with Parkinson’s disease; designed and published a pilot clinical trial to validate its accuracy against standard measurement techniques. Demo: vimeo.com/117463741

→ Developed an algorithm to identify temporal stability in commercial audio recordings; designed a web interface to facilitate tempo-based music playlist generation for recreational exercise or physical therapy. Details: https://goo.gl/NSW9wg

### Research Fellow in Neurology @ Beth Israel Deaconess Medical Center
Jan 2009 – Jan 2013 | Greater Boston Area
→ Authored two open-source toolboxes for exploratory data analyses of 3-D brain images. Over 750 downloads/registrations to date. Details: tools.robjellis.net

→ Published analyses exploring the impact of musical training on neurological changes in a cross-sectional sample of children and adults, and a longitudinal cohort of children measured over 3–5 years.


## Education
### PhD in Experimental Psychology
The Ohio State University

### Master of Arts - MA in Experimental Psychology
The Ohio State University

### BA in Psychology
University of Delaware


## Contact & Social
- LinkedIn: https://linkedin.com/in/robjellis
- Portfolio: https://robjellis.net

---
Source: https://flows.cv/rj
JSON Resume: https://flows.cv/rj/resume.json
Last updated: 2026-03-25