After twenty years in software engineering, I made the jump into data engineering and discovered that many of my instincts still worked — right up until they didn’t.

Some skills transferred cleanly: debugging, automation, systems thinking, and caring about reliability. But data engineering forced me to think differently about ownership, correctness, observability, and what it means for a system to be “done.” Building pipelines is not the same as building products, even when the tools and code look familiar.

In this talk, I’ll share the most important lessons from making that transition mid-career: the habits from software engineering that gave me a head start, the assumptions that caused the most pain, and the mindset shifts that helped everything click. If you’re curious about data engineering, considering a career pivot, or simply want a better mental model for how data systems work, this talk will give you an honest and practical view from someone who had to learn the differences the hard way.