Meridian receives thousands of scanned bills of lading per week and needs a consistent set of fields (shipper, consignee, container ID, port of discharge) pulled from every one into a table, with no open-ended question answering involved. Which Agent Bricks pattern fits this requirement with the least custom code?

Information Extraction is the correct fit because the requirement is a fixed, repeatable set of structured fields pulled from a high volume of similarly structured documents, which is exactly the managed pattern designed for that job, sparing the team custom extraction-chain code. Knowledge Assistant is the wrong choice because there is no open-ended question being asked here at all - the team wants the same fields every time, not free-form answers to varying planner questions. Multi-Agent Supervisor is the wrong choice because there is only one task and one document type in play; supervisor patterns exist to coordinate multiple specialized agents across different kinds of sub-tasks, which this single-purpose extraction job does not require. A hand-written regular-expression parser would struggle badly with the source format described - scanned, OCR'd documents with per-shipment variable layout are exactly the kind of noisy, non-fixed-position text that regex-based parsing fails on, whereas a model-backed extraction pattern tolerates that variability. Matching 'high-volume, fixed-schema, non-conversational' requirements to Information Extraction, rather than to Knowledge Assistant or a supervisor, is the core skill this objective tests.