
Abstract: Wouldn’t it be great if Python pandas worked with real-time, dynamic data? Or if publishing updating, derived data to apps and dashboards was easy? Deephaven Data Labs has an open solution exists today. Deephaven is a general-purpose data system built from the ground up to make working with real-time data easy -- on its own or in combination with large batch loads. Spun out of Wall Street trading, Deephaven provides intuitive, open, community tools and experiences for...
working with data from Kafka, redpanda, other message queues, and your custom streaming
sources;
performing table operations that update in real-time as new data arrives;
integrating your custom or 3rd-party Python libraries with real-time data;
applying data science, ML, and AI models on dynamic data;
exploring and interrogating;
creating real-time dashboards.
Session Outline:
This demonstration of Deephaven will show you how to:
1. Launch Deephaven from Jupyter and your Python IDE.
2. Access and manipulate Kafka and Parquet with just a few lines of code.
3. Filter, aggregate, join, select, and project ticking, updating tables in real-time.
4. Create and interact with table and plot widgets in web dashboards and Jupyter.
5. Integrate Deephaven with matplotlib, PyTorch, and other popular tools.
Bio: Pete spent more than two decades on Wall Street, growing, and running automated trading groups. In 2005, he was the founding CEO of Walleye Capital, a multi-billion-dollar quant fund that derives value at the intersection of real-time data and automated applications. In 2017, Pete and some engineers spun a proprietary data engine out of Walleye, forming an independent company called Deephaven Data Labs. Deephaven is an open-first software shop, delivering a real-time query engine, APIs, UIs, and integrations to the community via open projects designed for diverse teams. Deephaven complements streaming technologies and makes dynamic data easy and accessible.