Christopher Robert White

Experience

1. Lead Data Scientist · Nike

   Remote

   Jun 2023 - Present

   • Led technical deployment of a production 900K-variable weekly Mixed Integer Program (MIP) containerization model (gurobipy, Databricks) codifying 21 loading rules with rejection-driven feedback; scaled from pilot to 5 origin ports serving 17.5% of global inbound volume, lifting consolidator-bypass from 11% baseline to 45%+ and saving $2.8M annually.
   • Led ablation experimentation across 22 EMEA loading rules to isolate consolidator-bypass degradation on containerization rollout; identified ETA-zoning rules as dominant uplift lever (5-7pp), driving $236K in cost savings via post-launch parameter retuning.
   • Owned end-to-end a 413K-variable inventory allocation MIP (supply-demand matching, gurobipy on Databricks Apps), from stakeholder scoping through React/FastAPI front-end, clearing Nike marketplace demand across 23 DCs; delivered 157K-unit cross-territory divert ($2.5M) with $10M+ annualized avoided-cancellation savings for 27 analysts.
   • Owned a 3-person consulting squad end-to-end; mentored 8 data scientists, conducted 20+ hiring interviews, and led 30+ onboarding sessions.
   • Presented 10+ times at the enterprise Data Science Demo Series, sharing machine learning, experimentation, and supply chain optimization work with cross-functional audiences of data scientists, engineers, and business stakeholders.

2. Senior Data Scientist · Nike

   Remote

   Dec 2021 - Jun 2023

   • Architected a supply chain network Digital Twin (60K-variable MIP in Gurobi/OR-Tools) replacing Supply Chain Guru X, optimizing cost, lead time, risk, and emissions across 1.6B units via Pareto frontier, sensitivity, and scenario analysis; delivered $202M EBIT, $1M annual license savings, and 100M kg CO2e reduction in year one; shipped on AWS SageMaker via Jenkins CI/CD.
   • Trained CatBoost models on 180K Snowflake records for landed-cost pricing (FOB + duty) at SKU-factory granularity: 4.05% MAPE on FOB (56% below static weighted-average baseline), 19% error reduction on duty; monthly Airflow retrain, integrated into sourcing optimizers, served via Streamlit to 150+ designers.
   • Built 12 medallion-architecture ETL pipelines (Python, SQL) on AWS SageMaker and Databricks across 5 products, automating ingestion from Snowflake, Box, and Delta Tables; containerized in Docker and hardened with PyTest unit tests.

3. Senior Data Scientist · IBM

   Washington, DC

   Mar 2020 - Nov 2021

   • Built 2-stage MILP (Python/PuLP) combining multi-commodity network flow and bin-packing (materiel compatibility) for DLA munitions sourcing/distribution on Palantir Foundry with PySpark; cut planning from 1 week to 8 minutes across ~4K SKUs.
   • Engineered LSTM encoder-decoder (TensorFlow) forecasting 12-month backorders across 33K SKUs, using tslearn for segmentation and cluster-guided oversampling; benchmarked 6 architectures via walk-forward CV (0.89 F1, 0.94 recall); published DLA technical report.
   • Developed MLP/RNN classifiers (scikit-learn, TensorFlow) labeling Primary Failure Indicators from maintenance logs across 30+ failure modes at 85% accuracy; cut reliability curve development from 6 months to 4 weeks; adopted by the Air Force across 5 engine programs.
   • Developed bidirectional LSTM classifiers (TensorFlow, GloVe embeddings) fusing free-text and structured fields to classify security clearance applications at 0.91 recall and 0.89 F1 with SMOTE; SHAP force plots drove Army adoption.

4. Senior Technical Supply Chain Analyst · Accenture

   Arlington, VA

   Jan 2017 - Jul 2018

   • Enhanced DLA aviation SKU recommendations by building a hybrid retrieval pipeline (SAS, SQL) combining rule-based clustering (vendor, socioeconomic sourcing) with product keyword cosine similarity to extend candidate SKU sets; outputs pushed to a dashboard used by 300+ analysts, informing $1.2B+ contracting decisions.
   • Led development of a supply chain network dashboard in Qlik Sense for DLA distribution center performance monitoring; engineered supporting Python (Pandas, NumPy) and PostgreSQL data pipelines.

5. Program and Project Management Analyst · Accenture

   Arlington, VA

   Feb 2016 - Jan 2017

   • Supported HealthCare.gov modernization at CMS: monitored SDLC phase milestones, compiled weekly risk reports, tracked contract financials, and maintained vendor resource compliance across dev, QA, and infrastructure workstreams.