Data Science Meets Agentic AI: From Notebooks to Production (with Michael Kennedy)

​Software engineering discipline is becoming essential in data science and increasingly relevant in the age of agentic AI. In this episode, we talk with Michael Kennedy, host of Talk Python to Me and founder of Talk Python Training, about what it takes to build data-driven systems that run reliably in production.

​We’ll dig into the practical side of running Python systems at scale: managing dependencies, setting up lightweight CI/CD pipelines, adding observability, and tuning performance without unnecessary infrastructure. We’ll connect these ideas to the data science workflow, exploring how practices from software engineering can help AI teams ship faster and with more confidence.

​We explore which software practices matter most for data scientists, how to deploy Python systems without unnecessary complexity, and how agentic AI is reshaping development workflows. Together, we look at what’s working, where teams are running into problems, and how to write code that stays maintainable as projects scale.

​We’ll discuss:
🧩 The software engineering habits that make the biggest impact for data scientists
🛠️ Managing dependencies, deployment, and performance in production environments
📈 Reproducibility and testing in modern AI workflows
🤖 What’s actually working in agentic AI and why some teams are struggling

​If you’re building data systems that need to perform beyond the prototype stage, this conversation offers a practical look at how to bridge software engineering, data science, and agentic AI.

​​​​If you can’t make it, register and we’ll share the recording after.