Brief Bio
Christoph Sawade is the CTO and Managing Director at Fit Analytics Innovation GmbH. Before his current role, he was the Head of Engineering from 2022 until the end of 2023 at Fit Analytics under Snap Inc. Prior to joining Fit Analytics and Snap, Christoph spent three years as the Global Senior Director of Data at HelloFresh, where he led the data platform and data science departments. During his tenure, Christoph played a pivotal role in driving the adoption of a distributed data management architecture (Data Mesh). Prior to HelloFresh, Christoph worked for five years as the Director of Data Science at SoundCloud. In this capacity, he successfully designed and developed various large-scale data science products, most notably a robust audio fingerprinting system with audio and content-based classification capabilities. Christoph holds a PhD in Machine Learning and Computer Science ("Dr. rer. nat."), which he earned with the highest honors (summa cum laude), and a M.Sc. in computer science ("Diplom Informatik") from the Humboldt Universität zu Berlin. His doctoral research focused on the Active Evaluation of Predictive Models. For more details, see Christoph's LinkedIn profile.
Interests
Industry
Christoph's passion is to enable data products at scale including its organizational and technical challenges. In particular,
- Data as a Product: Data is used to solve real problems that maximise the experience of HelloFresh customers and the profitability of the business.
- Distributed Ownership: Product and Analytics Teams are empowered with tools, capability, and knowledge to own their data assets and develop their data products.
- Self-Serve Platform: The data platform reduces complexity by providing infrastructure, tools, and education on best practices, reducing the time to insights, and allow domain teams to focus on building data products.
Research
Christoph's research interests include theory and techniques in machine learning and statistics, in particular, for music classification / information retrieval and social networks. He has also worked on active learning and evaluation, learning under distribution shift, and structured learning in the context of scalable ranking of online ads.
Software
Check Christoph's Git repositories.
Publications
Industry
Clemence W. Chee and Christoph Sawade. HelloFresh Journey to the Data Mesh, 2021.
Özgür Demir, Josh Devins, Max Jakob, Janette Lehmann, Christoph Sawade, and Warren Winter. SoundCloud's Data Science Process, 2017.
General
Christoph Sawade. On the Pareto Principle, 2021.
Research
Matthias Bussas, Christoph Sawade, Tobias Scheffer, and Niels Landwehr. Varying-coefficient models for geospatial transfer learning. Machine Learning, doi:10.1007/s10994-017-5639-3, 2017.
Paul Prasse, Christoph Sawade, Niels Landwehr, and Tobias Scheffer. Learning to identify concise regular expressions that describe email campaigns. Journal of Machine Learning Research (16) 3687-3720, 2015.
Michael Großhans, Christoph Sawade, Tobias Scheffer, and Niels Landwehr. Joint Prediction of Topics in a URL Hierarchy. Proceedings of the 24th European Conference on Machine Learning (ECML-2014), Nancy, France, 2014.
Christoph Sawade. Active Evaluation of Predictive Models. PhD thesis, Universitätsverlag Potsdam, 2013.
Michael Großhans, Christoph Sawade, Michael Brückner, and Tobias Scheffer. Bayesian Games for Adversarial Regression Problems (with appendix). Proceedings of the 30th International Conference on Machine Learning (ICML-2013), Atlanta, USA, 2013. JMLR: W&CP volume 28.
Christoph Sawade, Steffen Bickel, Timo von Oertzen, Tobias Scheffer, and Niels Landwehr. Active Evaluation of Ranking Functions based on Graded Relevance (Extended Abstract). Proceedings of the 23rd International Joint Conference on Artificial Intelligence (IJCAI-2013), Invited Track on Best Papers from Sister Conferences, Beijing, China, 2013.
Christoph Sawade, Steffen Bickel, Timo von Oertzen, Tobias Scheffer, and Niels Landwehr. Active Evaluation of Ranking Functions based on Graded Relevance. Machine Learning Journal, 2013, 10.1007/s10994-013-5372-5.
Christoph Sawade, Niels Landwehr, and Tobias Scheffer. Active Comparison of Prediction Models (with appendix). Proceedings of the 26th Annual Conference on Neural Information Processing Systems (NIPS-2012), Lake Tahoe, USA, 2012.
Christoph Sawade, Steffen Bickel, Timo von Oertzen, Tobias Scheffer, and Niels Landwehr. Active Evaluation of Ranking Functions based on Graded Relevance. Proceedings of the 22nd European Conference on Machine Learning (ECML-2012), Bristol, England, 2012. Best Paper Award
Paul Prasse, Christoph Sawade, Niels Landwehr and Tobias Scheffer. Learning to Identify Regular Expressions that Describe Email Campaigns (with appendix). Proceedings of the 29th International Conference on Machine Learning (ICML-2012), Edinburgh, Scotland, 2012.
Christoph Sawade, Niels Landwehr, and Tobias Scheffer. Active Estimation of F-measures (with appendix). Proceedings of the 24th Annual Conference on Neural Information Processing Systems (NIPS-2010), Vancouver, Canada, 2010.
Christoph Sawade, Niels Landwehr, Steffen Bickel, and Tobias Scheffer. Active Risk Estimation. Proceedings of the 27th International Conference on Machine Learning (ICML-2010), Haifa, Israel, 2010.
Steffen Bickel, Christoph Sawade, Tobias Scheffer. Transfer Learning by Distribution Matching for Targeted Advertising. Proceedings of the 22th Annual Conference on Neural Information Processing Systems (NIPS-2008), Vancouver, Canada, 2008.
Reviewing
Christoph has reviewed for the major conferences and journals in the area of machine learning:
- Conference on Neural Information Processing Systems, 2013-2015
- International Conference on Machine Learning, 2012-2016
- Conference on Knowledge Discovery and Data Mining
- IEEE Transactions on Pattern Analysis and Machine Intelligence
- Machine Learning Journal, 2015, 2012.
Teaching
Several introductory and advanced courses on basic theory and techniques of machine learning. Please check out Christoph's academic website.