Data Scientist - Inventory Forecasting Optimization

ALMIRA ABDUREYIM

DATA SCIENTIST INVENTORY OPTIMIZATION & FORECASTING

***************@*****.*** Phone: 407-***-****

PROFESSIONAL SUMMARY

Data Scientist with 4+ years of experience using Python and SQL to develop forecasting, statistical, machine learning, and optimization solutions for operational decision-making. Hands-on experience building time-series forecasting and mixed-integer linear programming

(MILP) models for inventory replenishment, resource allocation, and store-level planning. Experienced in translating demand forecasts into actionable inventory recommendations while accounting for inventory availability, safety stock, and operational constraints. Skilled in model validation, performance monitoring, and communicating analytical recommendations to Operations and cross-functional stakeholders. SKILLS

• Programming & Data: Python, SQL, Pandas, NumPy, PySpark, Spark SQL

• Optimization & Operations Research: Mathematical Optimization, Mixed-Integer Linear Programming (MILP), Pyomo, HiGHS, Inventory Replenishment Optimization, Resource Allocation, Decision Variables, Objective Functions, Operational Constraints

• Forecasting & Machine Learning: Time-Series Forecasting, Demand Forecasting, XGBoost, Scikit-learn, Linear Regression, Logistic Regression, Decision Trees, Random Forest, Clustering, Statistical Analysis

• Inventory & Operational Analytics: Inventory Replenishment, Forecast Demand, Safety Stock, Inventory Availability, Shortage & Excess Inventory Analysis, Store Performance Analytics, KPI Analysis, Forecast Error Analysis

• Model Validation & Monitoring: Time-Based Backtesting, Model Evaluation, WAPE, MAE, Forecast-to-Actual Variance, Data-Quality Validation, Model Monitoring

• Reporting & Platforms: Tableau, Microsoft Excel, Looker, Google Sheets, Matplotlib, Google BigQuery, Databricks, Snowflake, AWS, Jupyter Notebook, Git, GitHub

PROFESSIONAL EXPERIENCE

Focused on forecasting, inventory optimization, and operational analytics across 108 stores, developing data-driven solutions to support inventory decisions, resource allocation, revenue planning, and store operations.

• Developed an inventory replenishment optimization model using Python, Pyomo, and HiGHS, formulating replenishment as a mixed-integer linear programming (MILP) problem with store-SKU replenishment quantities as decision variables and incorporating forecast demand, inventory availability, safety stock, and operational constraints.

• Integrated demand forecasts with optimization models to translate predicted demand into prescriptive inventory recommendations, identifying projected shortages, excess inventory, and replenishment opportunities to improve inventory allocation and utilization.

• Built an end-to-end 7-day forecasting workflow across 108 stores using Python, SQL, and XGBoost, including data preparation, exploratory analysis, leakage-safe feature engineering, time-based backtesting, and store-level model evaluation.

• Redesigned the forecasting approach into a global recursive XGBoost model, consolidating seven horizon-specific models into one shared model and reducing WAPE by 6% in backtesting while simplifying model deployment and monitoring.

• Translated forecasting and optimization outputs into actionable recommendations for Operations, supporting near-term planning, inventory decisions, resource allocation, and identification of operational risks.

• Implemented solution-validation and monitoring processes using WAPE, MAE, forecast-to-actual variance, and data-quality checks to evaluate deployed model performance and identify opportunities for model improvement.

• Investigated model deterioration and operational performance changes using store-level drivers including traffic, conversion, ATV, UPT, promotions, inventory availability, and revenue trends; initiated retraining when appropriate.

• Collaborated with Data Engineering and Operations on production handoff, reusable data-processing logic, data-quality validation, workflow documentation, deployment support, and continuous improvement of analytical solutions.

• Developed Tableau reporting to communicate forecast performance, operational KPIs, trends, and analytical recommendations to Operations stakeholders.

• Analyzed operational performance across 40+ retail locations using SQL, Python, and Google BigQuery, monitoring revenue, traffic, conversion rate, AOV, UPT, and other KPIs to identify trends, performance changes, and business drivers.

• Performed recurring and ad-hoc analysis for Operations, investigating unusual sales patterns, KPI changes, and cross-store differences and translating analytical findings into actionable recommendations for operational decision-making.

• Built and maintained analytical datasets and recurring reporting workflows in Google BigQuery, using Looker and Google Sheets to support performance monitoring and business analysis.

• Partnered with a Senior Data Scientist on statistical and machine-learning analysis, including EDA, behavioral feature engineering, clustering evaluation, and customer-segment profiling across approximately 720K customers.

• Partnered with Marketing and a Senior Data Scientist to evaluate an A/B test for a targeted retention campaign, measuring a 7% lift in repeat-purchase rate between treatment and control groups. Taught GIS and geospatial data analysis; guided hands-on projects involving data extraction, preprocessing, spatial-tabular integration, geodatabases, analytical workflows, visualization, and interpretation. EDUCATION

University of Central Florida, Orlando, FL

Graduate Program, Information Systems Aug 2025–Present Beijing Normal University, Beijing, China

Master’s Degree, Information Systems Sep 2013–June 2016 Beijing Normal University, Beijing, China

Bachelor of Science, Geographic Science Sep 2009–June 2013 DATA SCIENTIST TORY BURCH Remote Oct 2024–Present DATA ANALYST REFORMATION Remote Oct 2022–Sep 2024 GEOSPATIAL DATA LECTURER XINJIANG NORMAL UNIVERSITY China Jul 2016–Jul 2021

Contact this candidate