How to Fast-track Production Apps with Agentic Coding Tools

Coding agents like Claude Code are quickly becoming the default way developers build — but on their own, they lack the context and tool access they need to be effective in real enterprise environments. Without workspace-aware grounding, they generate plausible-looking code that doesn't actually ship.In this session, we show how the Databricks AI Dev Kit turns a general-purpose coding agent into a Databricks-native developer. By giving agents secure access to workspace context, data, and tools, they can make informed decisions and take real actions — creating assets, querying data, and updating workflows directly in your environment. We'll also demonstrate how MLflow brings observability to coding agents, so you can trace behavior, debug issues, and understand how decisions are made in production.You'll walk away with the blueprint to move from standalone coding agents to fully integrated, context-aware systems that actually get work done.