Agent policy for regulated pipelines
You lead a data-platform team running large Python/SQL pipelines that feed both internal analytics and a regulated financial-reporting dataset. The team wants to delegate more pipeline work to agents to keep up with demand. Draw the policy: where do you let agents generate pipeline code freely, where do they assist under control, and where is agent generation forbidden — given that some of this data is in scope for an audit and some isn't? Defend the call most people would push back on.
zone_pipeline_work(is_regulated: list[bool], changes_schema: list[bool], has_reconciliation: list[bool], has_lineage: list[bool]) → list[str][[false],[false],[false],[false]]out["agent_free"][[true],[false],[true],[true]]out["agent_assisted"][[true],[false],[true],[false]]out["human_only"]Treat the AI’s output as a draft to verify, not an answer to trust. Name the specific flaw and the input that triggers it, say how you’d catch it (tests, edge cases, reading critically), and how you’d re-prompt or decompose to get it right.
Vibe & agentic: describe the solution in plain language (or narrate it) and the coach grades your approach.