No two CRMs are alike. Your institution has years of customization built in: fields that only your team knows the meaning of, decision codes that reflect your specific enrollment logic, orphan records from a migration that never got fully cleaned up.
This is exactly the kind of detail that makes or breaks an AI deployment. A generic tool won't know what your custom stage codes mean, or that your "prospect" definition works differently from the industry default. It will give you answers that look right but aren't calibrated to your data.
Before your enrollment agent starts working, it learns your CRM.
The onboarding audit
When you first connect your enrollment agent, it runs a full audit of your CRM setup. It surfaces:
- Custom fields — what they appear to track and how they're being used across records
- Decision codes and stage logic — the institution-specific rules that shape how your funnel works
- Orphan records and data quality issues — gaps that could affect the reliability of any analysis
- Structural relationships — how records relate to each other in ways that aren't always obvious from the schema
From that audit, the agent builds a data dictionary — a structured map of your CRM's logic that it uses to interpret your data accurately going forward.
Why the data dictionary matters
The data dictionary does two things. First, it makes the agent's analysis reliable. When you ask "how are our transfer applicants performing this cycle?", the agent knows exactly which fields and codes to look at for your institution — not a generic approximation based on a different school's CRM configuration.
Second, it makes the agent's outputs shareable. The data dictionary isn't just internal to the agent — it's something your whole office can reference. It becomes a living document of how your CRM actually works, often surfacing institutional knowledge that was previously only in one person's head.
Built for your setup, not a template
The enrollment agent isn't pretrained on a single CRM schema and forced to fit yours. It adapts. Every institution starts with an agent that actually understands its data — the specific logic, the specific quirks, the specific history that makes your enrollment operation yours.
That's the foundation for everything that follows.
