Skip to content
Eren Ezgi headshot

Ezgi Eren

Ezgi Eren

Lead Scientist at PROS, Inc.

Ezgi Eren is a Lead Scientist in PROS, Inc.’s science and research team that focuses on cutting-edge innovation to help customers outperform in pricing, revenue management and sales effectiveness. Ezgi has led and contributed to several research and implementation projects during her tenure at PROS, including pricing and revenue management innovation for the car rental industry, which has been implemented in leading car rental companies with significant recognized revenue and margin gains globally. Ezgi’s work got selected to be presented in Grace Hopper Celebration as well as leading INFORMS conferences. More recently, Ezgi has developed an interest in an alternative approach to revenue management that also addresses challenges with uncertainty associated with the pandemic world and beyond. She has been leading a major innovation effort which resulted in an adaptive revenue management solution with less requirements compared to the traditional approach, utilizing Artificial Intelligence and Machine Learning. Ezgi holds a Ph.D. in Industrial Engineering from Texas A&M University.



Track: Revenue Management & Pricing

Are You Ready for the Post-Pandemic World? Adaptive Revenue Management Using Artificial Intelligence

Due to the disrupted supply chains, highly volatile supply and demand has become a norm for many industries since the start of the pandemic. Therefore, these industries need to be supply aware in their pricing. Although airline revenue management (RM) can address this problem, there is an increased need for more adaptive approaches as well as less rigidity in data requirements and assumptions, because many of these impacted industries have little experience with RM. The emerging “Data-Driven” approaches to RM could fill this gap. We present an innovative approach to RM that we call Direct Adaptive Neural Network Based RM (DiANNe). DiANNe is a non-conventional approach to RM as it skips demand forecasting and utilizes machine learning for generating prescriptions. It has an adaptive nature that handles volatility well. DiANNe not only makes RM accessible to more industries, it’s also robust to everchanging customer behavior. These properties make DiANNe more critical as we emerge from the pandemic.