Based on the candidate's resume, here are potential interview questions categorized by their experience and skills: Technical Architecture & AI Engineering

Agentic Orchestration: You built a custom AI orchestration layer for "Edward" using a streaming state-machine parser. Can you walk through how you handled partial or malformed LLM chunks to ensure the structured events remained valid?1 Agent State Management: In your "Agentic Chat" project, you used Postgres-backed checkpointing with LangGraph. What were the primary challenges in persisting state across long-running workflows, and how did you handle versioning of that state?2 RAG Optimization: Your RAG layer mentions "adaptive semantic caching." How do you determine the threshold for a cache hit, and how do you ensure the cached responses remain relevant as the underlying document store changes?2 Security in AI: You implemented SSRF and prompt-injection defenses. Can you describe your multi-layered approach to securing an application that executes AI-generated code or tool calls?2 Backend & Infrastructure

System Reliability: For "Edward," you designed isolated build sandboxes and a preview routing layer. How do you ensure complete process isolation between concurrent builds to prevent resource exhaustion or security breaches?3 Scaling Background Tasks: You utilized Redis and BullMQ for job queuing. How do you handle job failures or "zombie" processes in your build pipeline to ensure the system remains recoverable?4 Performance Monitoring: At Iterate AI, you built a real-time alerting system that cut resolution time by 50%. What specific metrics were you tracking, and how did you minimize "alert fatigue" for the engineering team?5 Frontend & Product Engineering