Fernwood Neighbourhood House delivers programs across food security, childcare, seniors' services, and community development. Our participant and program data currently sits across spreadsheets, paper intake forms, and several disconnected systems, which makes it difficult to understand who we serve and report to funders accurately.

We are preparing to move onto a shared CRM. Before that migration can happen, we need to know what data we actually hold and get it into a state worth migrating. We have already mapped which systems hold what. Building from that starting point, learners will document the data at the field level, then clean and consolidate it around a proposed data model.

Core tasks:

1. Document what fields each source captures and how records overlap
2. Identify duplicates, gaps, and data quality issues
3. Clean and consolidate into a single dataset
4. Propose a data model covering participants, households, programs, registrations, staff and volunteers
5. Document the assumptions and decision rules used throughout

Learners will deliver an audit report, a cleaned dataset, a proposed data model, and documentation of how they got there.

The audit report should be detailed enough that our CRM implementation partner can scope and price the migration from it. On the dataset, where a record is ambiguous we want it flagged and left as is rather than resolved on a best guess. A clear list of what needs staff review is more useful to us than a dataset that looks finished. The documentation of assumptions and decision rules matters as much to us as the cleaned data itself, since we will be maintaining this system long after the project ends.

Deliverables:

1. Audit report covering fields, overlaps, duplicates, and data quality issues
2. Cleaned and consolidated dataset, with ambiguous records flagged for staff review
3. Proposed data model
4. Documentation of assumptions and decision rules