HR & People

Payroll

Payroll runs: gross salary, pay period, payout bank details. Contains sensitive PII.

high riskpayroll_v1

Schema

Fields

6 canonical fields with multilingual hints and field-level validators. Any source column that resolves to one of these is mapped automatically by the cascade.

ColumnTypeRequiredHints / validators
employee_idstringyespersonalnummer · matricule · matricola · employee id · número de empleado
gross_salarynumberyesbruttolohn · salaire brut · stipendio lordo · gross salary · salario bruto
currencystringyeswährung · devise · valuta · moneda
pay_perioddatelohnperiode · période de paie · periodo di paga · pay period · período de pago
ibanstringiban · compte · konto · bank account · iban (CH, LI, DE, FR, IT, ES)
tax_codestringsteuercode · code fiscal · codice fiscale · tax code · código fiscal

Sample

Example data

6 rows, gross salary + payout IBAN. Fully synthetic — safe to share, commit, and run in CI.

Download the fixture and drop it into the Workbench, or POST it straight to the API to watch the cascade resolve a real, deliberately-messy header row.

payroll_sample.csv
template
payroll_v1

Try it

Use it

Map a file to this template with one call. The cascade resolves most columns for free on the deterministic layers; only the leftovers reach the metered LLM.

REST

POST the file with template=payroll_v1 to /v1/uploads. Schema-only mode sends only headers plus a few clamped sample rows; full-data mode runs row-level AI in-region under a BAA and spends a few tokens per map from your prepaid wallet.

curl
curl -X POST https://api.adaptivmapr.com/v1/uploads \
  -H "Authorization: Bearer $ADAPTIVMAPR_API_KEY" \
  -F "file=@payroll_sample.csv" \
  -F "template=payroll_v1"
Model Context Protocol

Drop AdaptivMapr into Cursor or Claude Desktop and call the same cascade as an MCP tool. Schema-only calls leave only column names and a few clamped samples — a data-minimization mode by design.

mcp
adaptivmapr.match_headers({
  template: "payroll_v1",
  headers: ["first_name", "last_name", /* ... */]
})
MCP install instructions →

Compliance

Full-data on this template

This is a high-risk template. Full-data mode routes the metered layer-5 call under a BAA and spends a few tokens per map from your prepaid wallet. Schema-only mode (headers + ≤3 sample rows, each clamped to 80 characters) is a data-minimization mode — only column names and a few samples ever leave your tenant. See pricing →

Ready when you are

Put payroll in production — without shipping raw records.

Spin up a key in minutes. Schema-only mapping runs on the free deterministic cascade; a $10 prepaid wallet covers the metered layer when you need it.

Schema-only is a data-minimization mode · a few tokens per map