Chris Umphlett Email & Phone Number

Lansing, Michigan Area

Jackson, Michigan Area

Created new Customer Experience Index (CXi) automated reporting. I brought together transactions and survey responses from our web, call center and field channels from disparate sources and created the data architecture in SQL Server for storing the data. I developed code to calculate CXi across different levels of aggregation, create backups and automatically detect data quality issues.Developing Machine Learning process to classify incoming phone caller intent utilizing Naive Bayes.Utilized various machine learning and statistical techniques to support analytical projects, such as k-modes classification, naive bayes, CART, anomaly detection and feature reduction.Evaluated department’s call center forecasting and provided recommended improvements in modeling methodology and benchmarking.Discovered and remediated many data quality issues, teaching the team best practices for ensuring quality and catching errors.Prevented $25k vendor investment for reporting solution by developing workaround.

Lansing, Michigan Area

• Learn and implement machine learning and statistical processes, including: anomaly detection, force-directed network visualization, logistic regression, exponential smoothing, unobserved component models, automatic dimension reduction and variable selection, text mining, decision trees and fuzzy matching.• Created automated forecasting system with semi-supervised machine learning umbrella to automate model parameterization. Forecasts are generated using Unobserved Component Modeling implemented with SAS PROC UCM. The parameterization of exponential smoothing components as well as additional exogenous predictors is achieved by iterating through each time series multiple times, using model output to determine the optimal state of each parameter. • Connect to and query directly from production DB2 databases using SAS SQL. We then create and manage large datasets (up to 100 million records).• Use discrete event simulation (R’s simmer package) to predict the effect of various staffing strategies on Operations KPIs.• While I have not served in a formal leadership role at Jackson I am a mentor to 2-3 interns per year and assist with recruitment of interns and full time team members. According to my managers I have earned the "currency of trust" from our senior management and am often consulted for my perspective on ad hoc analysis of senior management initiatives.

Kalamazoo, Michigan Area

Executed statistical analyses in SAS that contribute to Kellogg’s Customer Business Plan, including optimal feature price, display partners, and feature page performance.Created and/or adapted SAS programs that improve the data preparation process, such as a program that creates indexed seasonality coefficients for each product.Modeled the sensitivity of sales by product to gas price volatility, delivering a written report and creating a sensitivity diagram for each business unit.

Mt Pleasant, Michigan

Fall 2010 & 2011: Aided faculty members with research projectsSpring 2011: Taught an undergraduate Economics course

Master’S Degree, Economics

Bs Of Business Administration

Bachelor'S Degree, Business Administration And Management, General