Adrian obtained his PhD in discrete algorithms at the age of 20. He specializes in network science and modeling processes which involve graphs, time, and all things random. During a decade in academia, he ran projects on transportation systems, route planning, and logistics across Europe. He likes to experiment with data with a Python data science stack whenever he can. A big fan of competitive programming - and an even bigger fan of 24-hour contests - Adrian co-founded spoj.com, which has been used by about a million people to boost their programming skills. He happens to be the author of some of the most bizarre problem storylines you will find there. For research audiences, he has talked on topics ranging from synchronization in distributed systems to path-finding algorithms, including two Best Paper talks at major ACM conferences.
As one of the co-creators of Pathway, Adrian has spent the last two years shaping its development directions, contributing code, and being obsessed about usability.
Machine Learning models designed to work with streaming systems make decisions on new data points as they arrive. But there is a downside: model decisions can't be easily changed later when the model is updated with fresher data, user feedback, or freshly tuned hyperparameters. This is often a blocker for anomaly detection, recommender systems, process mining, and human-in-the-loop planning.
To deal with this, we'll demonstrate design patterns to easily express reactive data processing logic. We will use Pathway, a scalable data processing framework built around a Python programming interface. Pathway is battle-tested with operational data in enterprise, including graphs and event streams in real-world supply chains, and is now launching as open-core.
You will leave the talk with a thorough understanding of the practical engineering challenges behind reactive data processing with a Machine Learning angle to it, and the steps needed to overcome these challenges.