Round 1 – Coding Round: Rag Spring Challenge Coding Round 2 – Discussion Round with Rohan Gupta: He asked me about the issues I encountered while coding and told me to update the code with a table extraction approach from documents. I spent half my day researching and provided them with the best possible solution. Round 3 – with Shubhangi Misra (SDE1 - AI Engineer) The title of the round changed, and I was expecting questions related to my previous tasks — RAG, LLM-based projects, my problem-solving approach, and how I implemented solutions. But suddenly, she gave me a sample coding task in Python. 🤷🏻‍♀️ She mentioned they wanted someone who could write code manually without AI assistance and handle API integrations. I was expecting questions related to RAG, LLMs, or integration-based topics. Instead, she asked about a basic Python program. What was the point of all the effort I put into researching and experimenting with the RAG task then? What’s the point of integrating AI with the frontend and backend if they just test on simple Python scripts? They should ask standard questions like how to integrate APIs with LangChain and similar topics. Honestly, it felt like a complete waste of an hour.