Matthew Stirling Email & Phone Number

San Diego, CA, US

Encinitas, California, US

• Led team to develop and deploy 13 novel ML systems via A/B testing framework. Some notable systems:
• A pricing engine, comprised of 4 ensembled ML models and optimization, which raised profit by 15%.
• Multiple ML models to enhance the company's pooling optimization system which increased pooling rates by 43%.
• Recommender system which boosted engagement by 41% and offers by 66%.
• LLM research and implementation to reduce cost and increase win rates.Led experimentation culture at company:
• Team designed and conducted hundreds of A/B experiments.

Encinitas, California, US

Encinitas, California, US

• Researched and built two production ML models that collectively boosted profit 292%, pooling rate 43%, and 3x automation of covering loads.
• Created A/B testing framework for machine learning models.
• Developed data science roadmap via 9 POC ML models with business impact estimates.

La Jolla, CA, US

Primary Instructor role for a capstone course focused on optimal stopping theory. Coursework included optimal stopping algorithms, auction theory, classification, regression, model stacking, ensembles, inference, and multi-armed bandits.

US

Built xgBoost supervised regression click-rate prediction model which increased revenue 8%. Deployed with AWS, Kubernetes, and Docker.

US

Built NLP topic model (LDA) in Python to grow comprehension of advertisements from 14% to 99%.

Fort Worth, Texas, US

Team built the company’s main lending approval model using a GBM model in R.

Albuquerque, NM, US

Team lead of 24 data scientists with a role to coach, create proposals, drive assignment alignment, manage work on projects.

Albuquerque, NM, US

• Technical and project lead on 4 different multi-lab, multi-year machine learning projects focused in the NLP and text data spaces. Teams utilized Java and Python to implement models including LDA, mini-batch K-means.
• Led revision of recruiting to grow the team from 8 to 24 data scientists.

Albuquerque, NM, US

• Tech lead for novel Python & Gurobi production schedule optimization model that saved $40 million.
• Built a data mining tool that identified best-fit distributions for each manufacturing process to improve our team’s simulations.

Washington, D.C., US

• Managed team of 6 data scientists and operations research analysts.

Washington, D.C., US

• Co-Authored white paper which introduces technological opportunities for USPS.

Washington, D.C., US

• Developed a MILP optimization model in SAS identifying over $1 billion in savings within the USPS delivery network. Published and presented the model at the 2012 CRRI conference.
• Co-authored “The Untold Story of the ZIP Code” which was covered by The Atlantic and inspired Ireland's government to adopt a ZIP Code.

Washington, D.C., US

• Designed and developed a national facility decision model using MILP optimization in CPLEX. Model implemented in Network Rationalization project which saved $3 billion annually.
• Identified over $20 million in annual savings through numerous metro plan analyses.

Washington, D.C., US

• Awarded Lean Six Black Belt for completing 3 Lean Six Sigma projects that saved $1 million and improved delivery service from 87% to 95% success.

• Founder for 3 companies with successful exits: a marketing company with 500+ employees, an Internet retail company with 30 employees, and a service company with 8 employees

Binghamton, NY, US

• Modeled fluid mechanics and pressures of blood flow in the radial artery and its implications on an implanted BioMEMS device using MATLAB.

• Developed and managed business for 7 branch managers, 100+ employees, and 400+ customers
• Awarded International Manager of the Year in 1999 (500+ company managers)

Master Of Science, Operations Research

Bs, Industrial Engineering