machine learning

Predicting lead conversions to prioritize lead nurturing and identify opportunities for marketing to re-engage.

• Used in-production for 12+ months
• tidymodel ‘mapping’ to list-columns
  
• The data has been de-identified and is used with permission

big data visualization

Visualizing the current state of Electronic Health Record Interoperability.

• Top 10 Healthcare IT News’ article of 2018
  
• Synthesizes 3.2 million records using data.table
  
• The data is from two publicly available data sets

data-driven matching

Flexible / Fuzzy string matching on name and address using term frequency/inverse document frequency.

• Used in-production matching on 100K+ records
  
• Can be executed 100% in SQL to 100% in R
  
• The data is from two publicly available data sets

Literate R programming used throughout