10 min read
sbt 2 Is Here. Why Consider Mill?
sbt 2 improves how builds execute. Mill offers a reason to reconsider how we understand them, especially when a Flink application needs more than compilation and tests.
Systems Journal / Kirk Sweeney
Data solutions for a modern world.
Notes from building data platforms and software: practical patterns, sharp edges, and things worth learning in public.
Recent writing
All posts →6 min read
Scala 3 on Apache Flink: the Practical Path
How to write idiomatic Scala 3 Flink jobs with flink-scala-api, including type derivation, dependency setup, packaging, and the migration traps to avoid.
2 min read
The Agentic SDLC
A philosophy for software development where humans and agents share the same context, tools, workflows, and quality bar.
6 min read
Starting with (py)Spark
There has been a large shift in the industry in recent years to moving to python for data engineering. Spark one of the leading distributed data processing engines has great suppor