R Govys: Agentic Coding with GitHub Copilot demo
Details
In this session, Eli Holmes (NOAA) will give a live, “learn by watching” demo of agentic coding with GitHub Copilot. Rather than a traditional talk, this session will walk through building something in real time using an AI coding agent. You'll get to see how a scientific coder integrates this tool in her day-to-day coding tasks and combines it with GitHub code management.
Agentic coding is different from using a chatbot. These tools are designed for complex coding tasks—they can work across files, implement multi-step changes, and iteratively improve code. As you will see, these tools have layers of agents that plan, code, test, and review. The user’s role shifts toward clearly describing tasks, providing reproducible examples of problems or desired changes, and reviewing the agent's decision-making and re-directing if needed. These tools are already widely used in industry and academia and represent how software development is evolving. This session offers a practical look at what that workflow actually feels like for a scientific coder (not software developer). I will be working in Python or R.
Register Now - https://amstat.zoom.us/webinar/register/WN_i99LkogfQgKUeAA3mu4o3Q