Full Stack Data Scientist
Full Stack Data Scientist with 15+ years of experience building scalable data solutions and web applications. Currently serving as Director of Data Operations at Quote.com, where I architect serverless functions on Azure, develop partner integration platforms, and create real-time monitoring systems. My expertise spans the complete data pipeline from collection and analysis to deployment, with deep proficiency in R for statistical modeling and geospatial analysis, Node.js for serverless architecture, and cloud platforms including Azure and AWS. I've successfully built machine learning models for text classification, developed Chrome extensions for data automation, created interactive mapping applications, and designed call routing platforms with CRM integration. My work combines technical depth in data science with practical software development skills to deliver business-critical solutions.
Work Experience
Director of Data Operations
Ring2Media / Quote.com Westport, CT (Remote) October 2020 – Present- Created and deployed Node.js serverless functions to Azure for handling partner setup integrations and reporting
- Created and deployed quality control tooling on top of Ringba's APIs to handle orchestration and batch operations
- Processed geospatial files in R to determine serviceable partner footprints
- Maintained single source of truth to drive dashboards for business reporting
- Created a suite of rule based Slack alerts to stay ahead of downtimes and lower conversions
- Developed a partner agent portal with role-based permissions to operate as a real-time bidding integration with Ringba
Developer / Consultant
Marcus & Millichap Calabasas, CA (Remote) September 2020 – Present- Developed a Chrome extension to copy Salesforce and Sharepoint data to a proprietary web form
- Flagged and reviewed duplicate data based on touch points and string distances to improve on normalization techniques
Data Analyst
MSM/HRSA Washington D.C. (Remote) September 2016 – September 2018- Worked on the HRSA New Coding Schemes for Adverse Action and Malpractice Payment Reports Project
- Used string distance algorithms to create lookup tables of commonly misspelled values
- Built machine learning models (NLP with n-grams) to programmatically classify actions based on narratives for the reported action
- Outlined ways to determine if submitted reports were misfiled (initial/revision) based on field values and narratives
Data Analyst
MSM/HRSA Washington D.C. (Remote) September 2014 – September 2016- Worked as a Data Analyst on the Data Validation Study of the National Practitioner Data Bank (NPDB)
- Built statistical models to determined what upstream events can be addressed to limit downstream effects on data collection
- Determined correlations between missing fields through hierarchal clustering on a dissimilarity matrix
Data Scientist and Software Developer
MSM Data and Analytics Mesa, AZ (Remote) March 2008 – August 2020- Tracked Google AdWord performance based on inbound call volume and and A/B tested landing pages
- Developed a call routing platform with scheduling and a CRM drip system using the Tropo API
- Performed a time-critical rewrite of the platform to use Twilio after Tropo deprecated their API
- Created in-house automation, data collection, and data sourcing tools
- Built in-house data analysis tool similar to Tableau for insights into NPDB data
Skills
R Programming
DBI, dplyr, ggplot, h2o, lubridate, mlr, ODBC, Quanteda, sf, tidyverse
Javascript/Node.js
DuckDB, Express, pg, Sequalize, Socket.IO, swagger-stats, Tedious, tidy.js
Cloud Computing
Amazon EC2, Azure Function Apps, Docker, OpenFaaS, Portainer, Serverless Functions
General Development
DBeaver, Excel, Git, Markdown, Quarto, SFTP, SQL, Tableau, Tower, Webstorm
Projects
NiceRide Station Popularity
R, leaflet, sf
Interactive mapping of NiceRide destinations based on selected starting station
Visit ProjectFarmCow
GloVe, pgvector, Node.js
Uses vector embeddings (Global Vectors) to determine similarity of guesses to the weekly word
Visit Project