Jeffrey Herbstman, Ph.D.
603 W Grand Blvd., Detroit, MI 48216 jeffrey.herbstman@gmail.com (847) 951-3845
EXPERIENCE
Stryker, Kalamazoo, MI, 2021-2023
Sr. Data Scientist - Acute Care Research and Development
Created predictive risk models for in-patient falls, pressure injuries, and sepsis using Microsoft Azure, Jupyter Notebooks, Python
Managed data strategy, project development and designed real-time dashboards in Power BI and SQL to audit and understand streaming data from bed configurations in hospitals to prevent in-patient falls illustrating a 25% reduction in falls due to implementation of Stryker products
Analyzed data generated by R&D products to gain insights on usage, product errors to reduce errors 90%
OneMagnify, Detroit, MI, 2020-2021
Sr. Data Scientist - Customer Strategy and Insights Group
Built predictive in-market and likelihood to service models (random forest, gradient boosting) using SAS Enterprise Miner, Python (scikit-learn, XGBoost) for automotive marketing campaigns
Implemented a time varying parametric Customer Lifetime Value model in Python to identify the most valuable customers for marketing.
City of Detroit, Detroit, MI, 2017-2020
Sr. Data Scientist - Innovation Team - Mayor’s Office
Led quantitative efforts for human centered design internal consulting team tasked with solving some of the city’s most challenging problems
Built and refined a predictive model for a public health intervention using graph analysis and machine learning in R and Python based on demographic information, contact history, and the social interaction data
Collaborated with a broad range of stakeholders to aggregate, explore, and analyze data from a variety of sources. Developed metrics and analyzed success criteria for city programs around early childhood education, school evaluation, census outreach, and expungement of criminal records in Excel and R with dashboard visualizations in Tableau
Completed and communicated quantitative policy analysis for rollout, effects, and ROI of mayoral and city council initiatives, developed internal data governance policies, data warehousing strategies, analytics infrastructure solutions, and evaluated ambiguous requirements in dynamic, high-pressure environments
College of Creative Studies, Detroit, MI, 2017-2020
Adjunct Lecturer - Data Visualization - Graduate Studies
Designed and implemented an introductory course on data visualization in R
Instructed MFA students about visualization design principles and methods for clear and accurate communication of insights, data, and analytical methods
Teradata Corp., Detroit, MI, 2016-2017
Sr. Consultant, Data Science - Think Big Analytics
Collaborated with sales teams as a pre-sales consultant for the Aster Analytics Platform, the Teradata data discovery and advanced analytics tool. Worked with technical and non-technical customer business stakeholders to define project requirements, use cases, success metrics, and evaluation criteria for proof of concept work
Completed proof of concept solutions for modeling outcomes related to health and outpatient care solutions, the development of an time series IoT model to predict failure of machinery, smart vehicle sensor analysis, and customer journey analytics
Designed and delivered introductory data science presentations and advanced techniques training and demonstrations on a broad range of topics including predictive algorithms, time series methods, and text mining.
Served as an internal thought leader for projects around the Customer Journey, Connected Vehicle, Smart Cities, and Open Source integration with R and Python
Lochbridge, Detroit, MI, 2015-2016
Data Science Consultant - Team Lead
Led project management for team of analysts and data scientists for machine learning applications, anomaly detection, and analysis of time series sensor data for IoT applications with AWS IoT.
Built analytics engine and data pipeline with SparkML for processing of batch and real-time streaming data in Apache Spark on AWS using S3 and EC2.
Established a company center of knowledge for Data Science practices using R (caret) and Python (scikit-learn). Led internal data science trainings and brainstorming of use cases for data science projects.
Developed custom pricing models for a hospital system RFP response.
Marketing Associates/Magnify Analytics, Detroit, MI, 2013 - 2015
Sr. Business Intelligence Consultant - Manager
Onsite Data Science lead and manager for Voice of the Customer text analytics team for Fortune 100 company. Implemented text classification based on linguistic and frequency measures. Used machine learning to perform semi-supervised classification of text data using SAS, R. Extracted data for analysis using Hive. Developed tools for measuring classification metrics using SAS Enterprise Miner. Built key visualizations for measuring and comparing project progress using SAS Enterprise Guide, Tableau.
Generated and refined predictive models for customer behavior (likelihood to purchase, defect, credit risk, etc.) using R, SQL, and SAS. Build an imputation model in R to improve predictive accuracy of predictive modeling by 15%.
Led winning team for Connected Vehicle Analytics Challenge project integrating vehicle telemetry, and sensor data with customer information to enhance and build customer-centered predictive models using R, Python, QGIS, Hive, and Tableau.
Initiated learning sessions covering topics including Pig, Hive, and machine learning methods
PageKicker (Early Stage Startup), Ann Arbor, MI, 2013
Search and Natural Language Processing Technical Lead
Developed machine generated text summarization methods and analytical metrics for natural language processing startup specializing in automated generation of content using Python programming.
Built Python programs to call APIs to enhance automated information retrieval
University of Michigan Life Sciences Institute, Ann Arbor, MI, 2010-2013
National Research Service Award Postdoctoral Fellow
Imaged and analyzed large data sets comprising thousands of electron microscope images for structural study of protein complexes.
Discovered biologically significant sub-step in ribosome maturation through 3-D modeling, image clustering and classification using Python, C, and custom applications.
EDUCATION
2005 - 2010 University of Michigan
Ph.D., M.S. in Applied Physics Doctoral dissertation study of small scale laser damage for use in microfluidic biomedical devices. Excelled in a core curriculum of physics courses with elective work in statistics, biostatistics, machine learning. Awarded: Ph.D. Preliminary Exam - Pass with Distinction
2000 - 2004 Vanderbilt University
B.S. in Physics, Engineering Science Awarded: Summa Cum Laude, Engineering Science Program Award, VUSE Outstanding Scholar, Charles K. Bruce Honors Scholarship
PROFICIENCIES
Python
R
Spark
Jupyter Notebooks
ArcGIS Online
Linux/Unix Environment
AWS Cloud - EC2, S3, IoT
Microsoft Azure
Cloudera Hadoop (Hue, Hive, Impala) SAS (Enterprise Guide, Miner) Tableau
Power BI
Excel
Teradata
Aster
SQL HP Vertica (Certified)
MS Office Suite
AWARDS
Ford Connected Vehicle Analytics Challenge - First Place Award, People's Choice Award
Marketing Associates Bravo Zulu Award Internal recognition for outstanding performance
Ford Motor Co. Recognition Award, Golden Wrench Team Award Development and Execution of novel Voice of the Customer Project