CEO and Data Scientist @ Ehrenmüller
Julia is an open-minded mathematician, passionate about machine learning, artificial intelligence and data driven business models. Very recently, she started her own business offering consultancy and implementation in the area of predictive modeling and artificial intelligence. Before that, Julia worked as a Data Scientist at Sixt SE and at Robert Bosch GmbH, where she and her colleagues employed data mining and machine learning to increase efficiency in manufacturing. She started her career as a researcher in the field of extremal combinatorics and probabilistic graph theory at TU München and TU Hamburg, from where she received her PhD in 2016. In November 2017, Julia initiated the meet-up Data Science Stammtisch Allgäu, which offers regular talks and networking possibilities for data science enthusiasts in the Allgäu region.
Real-Time Recommendations based on Serverless Architecture
Whenever we use a search engine we would like the results being ordered in such a way that the first results are exactly what we we are looking for. However, this depends heavily on our personal interests and might therefore differ from user to user. To ensure an ideal user experience, recommendation engines that offer personalized, real-time recommendations are inevitable for many applications. In this talk we present a recommendation engine for an app that we have built and that will be launched in autumn. Recommendations are provided for each user individually based on his former behavior and interests. Our machine learning models are being retrained on a regular basis, where challenges like the cold start problem need special consideration. To ensure a scalable system, we perform our calculations serverless in AWS.