RAG Interview Prep/Project Review with Chetna Shahi

Whether you're preparing for an ML/AI interview or building a RAG system meant to run in production, this session is a focused walk through the parts that tend to cause trouble.

Most RAG material stops at the demo. This goes into what shows up later: version collisions retrieval struggles to resolve, silent truncation, chunking that splits the answer, evaluation that can mislead you. We talk through each and what usually helps.

Interview prep: we'll work through the kinds of questions that come up for senior RAG roles — retrieval vs generation failures, when RAG isn't the right tool, build vs buy, how you'd actually evaluate a system — and how to answer them clearly.

Project review: bring your setup and I'll share where I'd expect it to run into problems, and what I'd look at first.