Hilal Dönmez

• SPOC for DataMart Redesign: Acted as the primary liaison between designers and developers during the redesign of the DataMart, ensuring a seamless migration for all projects within my chapter. Led discussions to understand the impact of the redesign, provided feedback on requirements and limitations, and communicated updates to all stakeholders. Played a critical role in managing project dependencies, ensuring smooth integration, and supporting migration efforts across the chapter.
• Transaction Categorisation: As the primary point of contact and go-to person for this project, I led the development and deployment of a real-time and batch-based transaction categorization model to classify customer transactions into categories (e.g. Groceries, education), enhancing financial understanding and improving app engagement. I held full responsibility for each project phase, from model development and deployment to monitoring, alerting, and stakeholder management, ensuring continuous optimization and improvements in financial health and customer experience.
• Casual Inference Analysis for Digital Features at ING: Led an analysis project aimed at understanding the causal impact of transaction categorization and look-ahead features on customers' financial health and app engagement. I mentored junior data scientists throughout the project, guiding them in applying causal inference techniques, including propensity score matching, to gain insights into customer behavior. This analysis provided valuable recommendations to enhance the customer journey, improve financial literacy, and support cross-selling efforts by identifying key customer tendencies.
• RIDM and BLUES: Developed and improved income and expense detection models for private individuals to automate the extraction of financial data from transaction history, aiding in the decision-making process for consumer loans, credit cards, and overdrafts. By applying statistical approaches and rule-based methods, the models significantly reduced manual workload for the customer loyalty team during loan applications. Led stakeholder management, model enhancements, and was responsible for deployment and project planning, delivering an impactful solution that streamlined the lending process and improved operational efficiency.

• Customer Complaint Categorization & Risk Level Assignment: Led a data science team in developing a machine learning model to automate the categorization of customer complaints and assign risk levels, reducing manual triage for the customer solution center. Managed stakeholder relationships and ensured alignment with business objectives, delivering a solution that improved operational efficiency by automating key workflows.
• ATM Predictive Maintenance: Led the development of a predictive ATM maintenance model, enabling proactive issue detection and reducing downtime through improved scheduling and cross-functional collaboration.
• General Purpose Loan Default Prediction: Led the development of a machine learning model to predict loan defaults, allowing the company to proactively address high-risk loans and reduce defaults. Managed stakeholder expectations, coordinated project timelines, and collaborated with the risk management team to implement data-driven decision-making in the loan process.

Short term assignment for 6 months

Led the development and deployment of a machine learning model to detect income and expense patterns for private individuals, enhancing financial institutions' ability to make informed decisions on consumer loans, credit cards, and overdrafts.

Banking System Web Project using React and Java

Banking System Architecture