Istanbul, Istanbul, Türkiye
As the Head of Data Science at PttAVM, I lead the design and development of high-impact data products across multiple domains of e-commerce. With 10+ years of experience in forecasting, optimization, pricing, inventory analytics, retail data science, NLP, and modern ML/LLM practices, I focus on transforming end-to-end, complex business challenges into simple, reliable, and production-ready AI solutions. Throughout my career, I’ve worked on demand forecasting, replenishment, price optimization, inventory risk modeling, and AI-powered catalog management—delivering measurable improvements in both operational efficiency and commercial performance. My priority is to build high-ownership, high-trust teams that not only demonstrate technical excellence, but also operate with strategic clarity, strong execution discipline, and a mindset of continuous, fast-paced iteration—enabling them to deliver meaningful, scalable, and sustainable impact across the organization.
Replenishment (F4S) • Led end-to-end discussions with business units and managed the technical design and development of replenishment product features. • Defined system requirements, aligned cross-functional stakeholders, and guided implementation for scalable product delivery. Hierarchical Forecasting • Developed hierarchical time series forecasting models across multiple product aggregation levels to improve overall forecast accuracy. • Applied statistical and machine learning approaches to support top-down, bottom-up, and middle-out forecasting strategies. • Combined tree-based models with time series algorithms, achieving higher accuracy across complex product hierarchies. Warehouse Inventory Count Selection • Built a predictive model to prioritize SKUs and warehouse locations for inventory counting based on risk and operational impact. • Reduced operational costs by optimizing count frequency and selection strategy through data-driven recommendations. • Utilized XGBoost to identify high-risk SKUs and locations, enabling more efficient and targeted inventory auditing.
Dynamic Pricing Product • Led data science efforts within the Dynamic Pricing product team and guided end-to-end model development. • Designed an in-house dynamic pricing model, supported by a comprehensive literature review to determine optimal model structure. • Managed both technical and business aspects of the algorithm design, ensuring alignment, feasibility, and measurable commercial impact. Replenishment Product • Served as the lead data role within the Replenishment product team, collaborating closely with business stakeholders. • Led discussions on end-to-end business requirements and oversaw the technical design and development of replenishment system components. • Ensured that model architecture and operational workflows supported accuracy, scalability, and successful product adoption.
Transformation Data Team • Developed robust data analytics solutions aligned with business requirements, ensuring accuracy and scalability. • Delivered actionable, data-driven insights through advanced Tableau visualizations, supporting decision-making across operational teams.
Replenishment Product • Served as a Lead Data Analyst/Scientist and Business Advisor within the Replenishment product team, contributing to both technical and strategic decision-making. • Designed and developed the in-house Replenishment product, enabling the termination of the external vendor contract and reducing long-term operational dependency. • Led the end-to-end design of replenishment business processes and technical architecture, ensuring accuracy, scalability, and alignment with operational needs.
Replenishment Product • Worked as a Data Analyst/Scientist within the Replenishment product team, supporting both technical development and business alignment. • Designed a multi-algorithm demand forecasting system using statistical and machine learning approaches, including ARIMA, Holt-Winters, neural network–based forecasting, and regression models. • Developed a heuristic model selection framework to automatically identify the most accurate forecasting method across product groups. • Achieved at least 20% improvement in forecasting accuracy compared to previous implementations. • Provided analytical support for KPI dashboards and operational reporting. • The Replenishment product was successfully adopted by major retailers including ŞOK Marketler, Bizim Toptan, Ramstore (Kazakhstan & North Macedonia), Makro Market, and Sarıyer Market. Consultancy Projects / POCs Turkish Airlines — Passenger Load Factor (PLF) Forecasting • Built an Elastic Net Regression model for PLF forecasting, achieving a 30% improvement in accuracy over the previous model. • Provided forecasts used directly by Revenue Management teams for planning and optimization. • Contributed analytical insights and supported dashboard-based performance reporting. OPET — Fuel Demand Forecasting (Station Level) • Developed a Stepwise Regression model to forecast fuel demand at station level, supporting campaign and operational decision-making. • Delivered model findings and insights through structured analysis and visualizations. İGDAŞ — Natural Gas Demand Forecasting • Built a hybrid forecasting model combining ARIMA and Stepwise Regression to improve natural gas consumption predictions. • Presented analytical outputs and model insights through detailed reporting and visualization.