Benjamin Greve

Automatically providing tens of millions of customers with a fully personalized selection from hundreds of offers - every day. Optimizing offer selection based on past purchasing behavior in order to maximize revenue, redemption rates and customer satisfaction. Building robust and efficient data pipelines to calculate hundreds of statistics for billions of customer-offer combinations. Developing, training, evaluating and deploying dozens of predictive models. Reducing complexity and interdependencies.

Developing and implementing a predictive model to identify poorly performing job ads in the website's job market section and automatically launch targeted counter measures. Implementing a web-based software framework for product managers to manage content exports to affiliates. Performing data integration of tracking data and webserver logs on a CDH big data cluster and running analyses on the cluster to answer business-side questions.

Forensic data analysis. Detecting anomalies in SAP accounting data. Text mining in large amounts of semi-structured text documents using Java and Lucene (financial services). Automatic recognition of repeating patterns in categorical time series data to identify fraudulent behaviour (retail industry). Developing and implementing algorithms to detect fraudulent behavioural patterns in online shop orders and activities.

Modelling and visualization of complex networks. Python- and Java-based plugin programming for Tulip (LaBRI) and Gephi. Development and implementation of a supervised text classification algorithm for bank transactions with R.

Wahrscheinlichkeitstheorie und Statistik