MongoDB Data Model Plan

PROJECT CONTEXT:
I am building on MongoDB and want a data model designed around my app's actual access patterns.

GOAL:
Produce a query-driven MongoDB model: collections, embed/reference decisions, indexes, and validation.

REQUIREMENTS:
- Access patterns listed before modeling
- Embed vs reference decided per relationship with reasoning
- Indexes matched to the query patterns
- Validation strategy (app-level or JSON Schema)
- The model tested against the real queries

CONSTRAINTS:
- Model for the queries, not for relational habits
- Avoid unbounded embedded arrays

DEFINITION OF DONE:
- Every app query maps cleanly to the model
- Indexes exist for every frequent query
- Ownership fields enable per-user scoping
- Validation rejects malformed documents

COMMON FAILURES TO AVOID:
- Modeling like SQL with joins everywhere
- Unbounded arrays that grow forever inside documents
- No indexes, causing collection scans on every list
- No ownership field, making user scoping impossible