Hi @Louis_Frolio . Nice question 😉

Two years ago, I was already using AI tools like ChatGPT and Copilot, but mostly as thinking partners to explore ideas and reason through the best solution. Before that, everything was manual: I spent hours searching through documentation, forums, and error messages with no real support.

Today, tools like GitHub Copilot and Claude feel like trusted developer teammates. They help me move faster, reduce repetitive work, and focus more on design and problem-solving. This is also true in my Databricks work, where AI and Claude skills are making development much smoother. In fact, these days I’m building my own skills to speed up the creation of agents with the Mosaic AI framework combined with LangGraph.

One of the biggest wins for me is documentation and README generation, which used to be heavy, boring, and very time-consuming. At the same time, what I’m realizing more and more is that experience still matters a lot: AI can accelerate execution, but hands-on knowledge is what helps you catch mistakes, challenge weak outputs, and avoid problems the model may not fully understand.