Building a Context-Aware Agentic Coding Platform for Airflow: Local Dev to Production

Generic AI coding assistants like Cursor and Claude Code are powerful, but they struggle with proprietary infrastructures. At Wix, managing 7,500 active DAGs across 120 Data Engineers, we found that standard AI tools lacked the context to be truly effective - they didn’t know our custom operators, DWH modeling patterns, or strict governance rules.

In this session, we’ll show how we built a context-aware agentic coding platform that bridges this gap across the full pipeline lifecycle. You will learn how we enabled our coding agents to:

• Write compliant DAGs using agent rules, skills, and plugins that encode production standards, naming conventions, and modeling patterns - distributed across the organization via our internal AICM tool
• Debug locally using skills that tap directly into running Airflow containers, and surface logs and fix failures without leaving your IDE
• Resolve production incidents autonomously via AirBot, our Slack-native AI agent that catches pipeline failures, pulls the DAG’s logs, and suggests fixes before the on-call even sees the alert

Whether you’re trying to optimize your team’s workflows or curious how far coding agents can go today, join us in this exciting talk.