Jeffrey Allard

About Jeffrey Allard

Data Scientist specializing in applying machine learning and statistical modeling to multi-channel marketing, campaign optimization, business and customer intelligence and market research. Passionate about open source technology and learning new techniques and methodologies in machine learning / predictive modeling. Extensive experience in bringing together and analyzing customer behavioral, transactional, survey and third party data, building models, designing experiments and driving test-and-learn methodologies across various industries and channels off-line (direct mail, telemarketing) and on-line (e-commerce, email marketing, search engines). Unique experience set that includes business / strategic responsibility as well as advanced analytics, allowing me to not only have the technical knowledge to apply cutting edge methodologies but to also understand the needs of the business and how analytics can help drive growth and financial performance. Technical skills leveraged to help answer business and research questions include: • Machine learning / Deep learning • Uplift / CATE models • Statistical modeling / inference / hypothesis testing • Design of experiments / Power Analysis • Causal inference / Intervention analysis • NLP • Clustering / Segmentation / Topic modeling • Recommendation Engines • Multi-arm bandits / Contextual bandits • Time Series Forecasting (Arima and DL) • Attribution and media mix Software / Programming Languages: • Python, R, SQL, PySpark, SAS • Python ML / DL libraries include: pytorch, TF / keras, scikit-learn, lightgbm, catboost, xgboost, mllib, vowpal wabbit, pandas, numpy, pymc, prophet, genism, hugging face, matplotlib, lifelines, implicit, shap • Cloud : AWS / Azure / GCP • VS Code • MS Office

Experience

1. Sr Staff Data Scientist

   Vericast

   Aug 2021 — Present

   Technical data science SME (digital, contextual, financial services) • Lead data scientist on marketing platform development including propensity models, LTV, uplift models, product recommendations using FM / LTR, bandit-based treatment optimization, multi-touch attribution using RNNs • Developing new approaches for digital marketing : CTR prediction and contextual advertising

Education

• Central Connecticut State University

  MS, Applied Data Mining, 3.9

  2008 — 2013

• Aquinas College - Grand Rapids

  BA, Economics (Honors)

  1995 — 1997