MARIYAM FAWAZ

• Led the development and execution of a global AI/ML forecasting strategy across multi-brand, multi-channel operations.
• Built and managed a team of 10 data scientists in Budapest, Hungary to deliver scalable, accurate, and explainable demand forecasts.
• Improved forecast accuracy and stability for in-line products, new launches, and promotions.
• Integrated commercial insights and regional dynamics into model design to support strategic planning.
• Partnered with cross-functional teams to embed forecasting models into enterprise systems.
• Fostered innovation and excellence through mentorship and adoption of cutting-edge AI/ML techniques.

Integrated Business Planning, Forecasting, Pricing and Optimization

• Operationalized 5,000+ ML-based demand forecasting models, improving omni-channel revenue by $38M and supporting Integrated Business Planning (IBP) with short-term forecasts for inventory trading and long-term forecasts for optimal buys.
• Led development of machine learning–driven Inventory Trading Tool and optimization engine (Python, PuLP) to reallocate tradeable inventory across 900+ wholesale accounts, maximizing utilization and improving trading opportunity by 5%.
• Built New Product Forecasting and Size Curve Optimization models using computer vision and ML, unlocking $14M in incremental sales and $35M in cost savings by reducing out-of-stocks and excess buys.
• Developed topline revenue forecasting and category level forecasting models to support category planning team, to get plan level right and used as causal for improving accuracy of mix.
• Co-developed AI/ML-powered in-season markdown optimization tool that dynamically adjusted pricing based on sell-through performance, contributing to $70M revenue boost by reducing deep clearance and inventory buildup.
• Created algorithms to estimate lost sales due to out-of-stock
• Built predictive ML workflow to forecast vendor delivery delays, improving on-time performance and reducing markdowns resulting in $60M in cost savings.

• Developed framework to forecast outbound shipments from DC to optimize container space, support labor and capacity planning

Integrated Marketing and Customer Analytics

• Led development of marketing performance frameworks for CMOs across multiple brands, building KPIs (CAC, ROAS, CPC, CLV, AOV), Multi-Touch Attribution (MTA), and Media Mix Models (MMM), resulting in $26M in incremental revenue.
• Built churn prediction models to identify customers likely to churn within 3 months, enabling retention strategies that drove a 3% revenue lift.
• Directed digital marketing variable spend program using causal inference and forecasting, revealing $2.5 return for every $1 of additional spend.
• Developed customer segmentation (lookalike) and propensity models, integrated into CDP (BlueConic) to power targeted campaigns and personalization, achieving 7% lift in ROI.
• Built customer cohorts across marketing channels to inform digital strategy and improve cross-channel targeting.
• Developed event-based and seasonal models to drive audience selection for catalog marketing, improving targeting precision and campaign efficiency.
• Created Revenue Sensitivity Modeling framework to assess impact of catalog marketing budgets on demand, supporting strategic investment decisions.
• Built uplift models (Double Robust Learner) to measure true incrementality of catalog marketing efforts.
• Extensive experience with Funnel, Datorama, BlueConic, Experian, CRM, Loyalty Data, Adobe, Bluecore, Zeta, and GA4 for campaign optimization and measurement.

Generative AI and ML Engineering

• Led migration of legacy analytics systems from SAS to AWS, modernizing data infrastructure and enabling scalable, cloud-native solutions.
• Developed and deployed a front-end application on AWS EC2, leveraging Stability AI’s SDXL model to generate product images from 2D sketches—accelerating product lead times and reducing design costs.
• Built a Retrieval-Augmented Generation (RAG) system to automate product description generation for an e-commerce platform, using product attributes and consumer personas to create a vector store—resulting in more accurate, engaging content and improved content generation efficiency.
• Applied RAG architecture for customer intent classification, retrieving relevant historical queries and support documents to enhance classification accuracy and improve customer experience through context-aware AI responses.

• Delivered customer segmentation, marketing analytics, and dashboard solutions for Bed Bath & Beyond, Academy Sports, and Carolinas HealthCare System—leveraging registry, sales, and campaign data to generate insights, evaluate ROI, and support targeted marketing using KPIs such as sales deciles, conversion rates, lead quality, and budget efficiency.

• NLP: Reduced suspicious medical provider referral to Claim Investigative Unit by 30% and refined business rules by implementing text mining (topic modeling) on claim investigation comments for possibly fraud/suspicious medical claims
• Medical Provider Fraud Model Enhancement: Improved accuracy by 7% of Medical Provider Fraud Identification model by revising feature engineering and introducing new features into model. Revised model data preparation code by introducing SAS hash lookups to reduce ETL processing time from 4 hours to 30 minutes

• Built a logistic regression model for prepaid churn prediction (Telecom), driving 2% incremental revenue and supporting win-back and retention strategies.
• Segmented retail banking customers using CHAID to identify channel migration opportunities, scoring branches for pilot targeting and goal setting.
• Improved ETL efficiency and reporting by streamlining data exploration and automating ad-hoc reporting tasks.