Large Language Models (LLMs) present a transformative opportunity to address longstanding challenges in computing education, such as scalability. This paper presents a web-based, conversational AI extension to an existing LLM-enhanced C/C++ compiler to generate pedagogically sound programming error explanations. Our new tool, DCC Sidekick, retains integration with the compiler, allowing students to see their code, compile- and run-time error messages, and stack frames alongside a conversational AI interface. This approach utilises compiler error context to improve error explanations, and provides a seamless user experience. We present quantitative analyses of the tool’s usage and engagement patterns in a large Australian CS1 course. In the first seven weeks of use, 959 students initiated 11,222 DCC Sidekick sessions, generating 17,982 error explanations. More than half of all conversations occur outside of business hours, highlighting the value of these always-available tools. Early results indicate strong adoption of conversational AI debugging tools, demonstrating scalability in supporting large CS1 courses. We share implementation details and lessons learned, offering guidance to educators considering integrating AI tools with pedagogical guardrails.