Do you really need cloud providers to run a large-scale data platform?

In this talk, Vakaris Baškirov will answer that question through the case study of Humbility, a high-frequency trading firm that relies on massive datasets to develop and evaluate trading strategies.

He will walk through the key decisions that led Humbility to build a petabyte-scale data platform on premises using open-source technologies such as Apache Spark, Apache Iceberg, and ClickHouse. He will explore the architecture, the trade-offs, and the challenges encountered along the way.

By the end of the talk, attendees will have a better understanding of the pros and cons of running a large-scale data platform on premises, the lessons learned, and whether this approach could be the right choice for their organizations.