Importing data

Most businesses arrive with history in a spreadsheet or an export from another tool. The CSV importer is built to move that history in without losing the properties that make Wavetable trustworthy.

How it works

1. Upload. Drop a CSV export from your current tool.
2. Map. Match your columns to entity fields. Wavetable suggests a mapping; you confirm or correct it. Nothing imports until you have seen the mapping.
3. Import. Each row becomes a set of claim proposals, not raw rows. They carry the import as their source, flow through adjudication, and land in projections like any other accepted fact.

Why claims and not rows

Because imports are where duplicate and stale data usually sneaks in. Since imported facts are claims with provenance:

• identity resolution sees them and proposes merges against people who already exist from Gmail or Stripe, instead of silently creating duplicates,
• a bad import is recoverable: claims can be superseded and projections rebuilt,
• you can always answer “where did this value come from?” even years later. The receipt says: this CSV, this row, this day.

Practical notes

• Re-importing an updated file proposes new claims rather than overwriting; conflicts surface with both sources visible.
• Dates, amounts, and emails are validated at the boundary; rows that fail validation are reported, not silently dropped.