Hi all,

We are excited to invite you to this exciting virtual meetup with a very interesting presentation from Humn.ai.

Abstract: A practical overview of Humn.ai’s empirical dissertation around how a real-time high-frequency data stream can be best joined for enrichment purposes with a slow changing stream that maintains state. This walkthrough covers the use of different Flink APIs, such as the DataStream and the Table APIs, using techniques such as Temporal Tables, Broadcasts and Heartbeats and includes pros and cons of the different options to help the audience make the best choice.

Info about the presenter: HUMN.ai are a well-funded and well-connected deep-tech startup that operates in the insur-tech industry. We have built a platform to perform dynamic risk assessment of the world around us, and how it impacts our customers, which are fleets of cars typically (but not exclusively) in the ride-hail business. We run streaming analytics using telematics data from the cars down to sub-second granularity as well as hundreds of other data sources, and use machine learning to come up with a risk model and a dynamic pricing engine, on top of which we build our insurance products.