Here's a refined version you can use for interview feedback or LinkedIn notes: The process started with an HR screening round focused on background verification, previous experience, and an overview of my technical skill set. The first technical round was with a Senior Software Engineer (5+ years of experience). Based on the job description and initial discussions, I expected questions around GenAI fundamentals, Agentic AI concepts, and some DSA. However, the discussion was almost entirely focused on AI architecture and AI system design and lasted approximately one hour over Zoom. While I was generally prepared for GenAI and Agentic AI topics, I realized there were a few areas where my depth of knowledge could be improved: RAG evaluation techniques and metrics Agent evaluation frameworks Context management and memory strategies in AI agents Fine-tuning approaches and trade-offs Advanced LLM application architecture patterns My takeaway from the interview is that the team is looking for someone with deep expertise in AI/LLM systems rather than a traditional Software Engineer who has incorporated AI into existing applications. Overall, it was a good learning experience and helped me identify the areas I need to strengthen to move from being an SDE with GenAI experience to a specialist capable of designing and evaluating production-grade AI systems. Key Topics to Strengthen RAG Evaluation Context Precision Context Recall Faithfulness Answer Relevancy RAGAS, DeepEval Agent Evaluation Task Success Rate Tool Selection Accuracy Multi-step Reasoning Evaluation Agent Benchmarking Frameworks Agent Memory & Context Short-term vs Long-term Memory Vector Memory Conversation Summarization Knowledge Graph-based Memory Fine-Tuning Full Fine-Tuning LoRA / QLoRA PEFT Techniques When to Use RAG vs Fine-Tuning AI System Design Multi-Agent Architectures RAG Pipelines Evaluation and Monitoring Guardrails and Safety Cost, Latency, and Scalability Trade-offs This interview suggests they are hiring for an AI Engineer / LLM Systems Engineer profile rather than a conventional Backend/SDE profile with GenAI exposure.