High-volume customer support operations face persistent challenges: inconsistent adherence to Standard Operating Procedures (SOPs), continuous human agent ramp-up cycles driven by attrition, and high Average Handling Time (AHT) due to repetitive context reconstruction across chat history and backend order views. We present AgentChat Copilot, a human-in-the-loop LLM system deployed at Flipkart, one of the largest e-commerce platforms in India. The system drafts SOP-grounded responses using structured order context and retrieved policy knowledge, while human agents remain responsible for reviewing/editing messages and performing all operational actions.