Databricks
How UC Berkeley researchers gave away Apache Spark for free and then built a billion dollar business on top of what they open-sourced.
In twenty thirteen, a group of researchers at UC Berkeley built Apache Spark — a data processing engine that ran one hundred times faster than Hadoop for most workloads. Then they gave it away for free. Open-sourced the entire codebase. Within two years, over a thousand companies were running Spark. It became the default tool for big data processing across the entire industry. The founders then started Databricks to sell the commercial version. Same technology, but managed, optimized, and supported. Enterprise customers pay five hundred thousand to ten million dollars annually for what is essentially a premium layer on top of free software. The open-source version creates the dependency.
Watch the full reel free on MoonReelz — moonreelz.com