AI and analytics leader · Salesforce Data Cloud
For more than twenty years I have helped organizations in healthcare, higher education, e-commerce and finance turn scattered data into models and decisions people can act on, and explained the results plainly to the executives and clinicians who use them.
A full scholarship that made a Reed College education possible, an Amazon Inventor Award, and research published before graduation.
Opportunity Program, Class of 1997 · Salem, Oregon
Chosen through a competitive selection and interview process for self-initiative, ability to learn and leadership potential. The full scholarship carried her through Reed College.
Amazon · Seller Recommendations
Awarded as an Amazon inventor for work on recommendation models that predicted merchant sales opportunities, product demand and fulfillment placement.
Choices magazine · Second Quarter 2000
Co-author, “Can Open Spaces Be Self-Financing?” Choices 15, no. 2 (2000): 1–3.
How nearby parks and natural areas raise property values and tax revenue in the Portland metro region. Elisabeth researched the topic in her junior year in Professor Noelwah Netusil's course on nonmarket valuation, and the paper was published by the Agricultural and Applied Economics Association under a former name.
At Antioch University I built the institution's customer data platform on Salesforce, from the first API ingestion to measured marketing results. I now do the same work for clients as a data science consultant at Jax Consulting.
Siloed systems
One unified profile
Predictive models
Activation
Estimated lift in submitted applications associated with the segmentation and A/B testing program.
Ingestion, identity and the data model are designed before the first dashboard, so the profile stays trustworthy as new sources arrive.
Einstein Studio models built for a real decision, then fed back into the segments marketing teams use.
Segments launched with a test plan, so the organization could see what worked and what to stop doing.
Four projects that show the range: higher education, population health, workforce planning and e-commerce.
Enrollment, marketing and student-support data lived in separate systems, and critical jobs needed manual onsite restarts. I moved those jobs to a remote virtual machine environment, built cross-silo pipelines in SQL and Tableau Prep, and configured Salesforce Data Cloud as the single customer view.
I integrated the Johns Hopkins Adjusted Clinical Groups model into a SQL Server pipeline with crosswalk tables and a queue-based, multi-threaded framework, producing multi-morbidity segments and risk scores for health plans and care managers. I then worked with payer and provider clients to apply those scores to care decisions.
Drawing on my training as an economist, I led tuition elasticity modeling and built regional supply-and-demand models that show where workforce gaps and market need are largest.
On the Seller Recommendations team I partnered with engineers on machine learning models that predicted merchant sales opportunities, product demand and fulfillment placement. Later roles at Avvo and Monster extended this to propensity-to-buy and sales attribution models.
Equally at home designing a data model, tuning a predictive model, and presenting the findings to a leadership team.
Analytics roles across consulting, higher education, healthcare, technology and finance.
AI, analytics and customer data work across complex CRM and cloud environments: Salesforce Data Cloud, Einstein Studio, Marketing Cloud, unified profiles and identity resolution, API ingestion design and production segmentation.
Built the analytics, data and delivery foundation for Enrollment Management Operations, including Salesforce Data Cloud, the university's first machine learning models, and the introduction of Agile Scrum, Jira and Confluence.
Analytics engineering with Python, Snowflake, PostgreSQL, dynamic SQL and Tableau.
Integrated the Johns Hopkins ACG risk-adjustment model into a SQL Server pipeline, cutting run times from days to hours, and worked with payer and provider clients to turn risk scores into care-management strategies. DataSmart was acquired by TCS Healthcare, now Chordline Health.
Python and machine learning for sales analytics, propensity and attribution.
Seller Recommendations: models predicting merchant opportunities, product demand and fulfillment placement. Amazon Inventor Award.
Econometric models forecasting consumer response to online offers, automated revenue reporting, and lead-inventory optimization.
Manager of Data and Analytics at TradeKing.com and Scorecard Analyst at Morgan Stanley, both in SAS.
Portland, Oregon · 2000
Bachelor of Arts in Economics, completed on a full Ford Family Foundation scholarship. Published research with the economics faculty before graduating, and wrote a senior thesis on the profitability of a USBancorp acquisition.
Agile and analytics
Certified Scrum Master. SAS Programming, Levels 1 and 2.
I'm glad to discuss analytics leadership roles, Salesforce Data Cloud work, or consulting engagements.