What records can MirrorFirm generate for an intake workflow?

A release for intake can contain the records an agent reads and writes, generated to be consistent with one another.

The core is a set of **cases**, each with a submission, a channel, a received time and a lifecycle of status changes. Around each case sit **parties** (applicants, existing customers, brokers or referrers), **documents** with type, legibility and dates, and **notes** written by staff roles at handover.

Policies are generated as well. A completeness policy can list required fields and documents by case type, with tolerances and exception rules. A routing policy can map case attributes to queues, including tie-breaks for cases that match more than one rule.

What makes the set useful is relationships. A duplicate submission points to an earlier case. An amendment references the field it changes. A document attached to the wrong case is detectable because its party does not match.

Each case can also carry an expected outcome: the correct status, queue, reason codes and outgoing message type. That is what turns a dataset into something an agent can be graded against.

All of this is generated, not copied from real business records. The generation is informed by direct observation of how real businesses operate, but a buyer should still inspect samples for grounding, and for any unintended resemblance to real people or organisations.