Revolut Business

Use the Revolut Business source for treasury and reconciliation reporting. Each run lands a daily snapshot of account balances and upserts every transaction and transaction leg in your sync window, so status changes (for example pending to completed) update the row you already have.

Counterparty names, references, and card details never land. Transactions arrive as amounts, currencies, states, and IDs.

Before you begin

Set up API access in Revolut Business. In the Revolut Business web app, go to Settings → APIs → Business API and add a certificate. You generate an RSA key pair, upload the public key (X.509 certificate), and set an OAuth redirect URI. Revolut gives you a client ID.
Authorise the app once to get a refresh token. Follow Revolut's consent flow for your client ID and exchange the authorisation code for tokens. Keep the refresh token and the private key you generated.
Note your issuer domain. This is the domain of the redirect URI you registered (for example example.com). Skipprd puts it in the signed client assertion it sends to Revolut.
Pick a destination that supports how Revolut lands. Transaction tables upsert by ID and the balance snapshot replaces each day's partition. Only Athena Iceberg and SkipprLake support both.

To try the integration first, create the certificate in a Revolut Business sandbox account and set api_base to https://sandbox-b2b.revolut.com/api/1.0.

Configure

Export the credentials so the config can reference them without storing them:

bash

export REVOLUT_REFRESH_TOKEN="oa_prod_..."
export REVOLUT_PRIVATE_KEY_PEM="$(cat privatecert.pem)"

Add the source and wire it into a pipeline. These examples assume you already have a destination named lake.

PythonCLIYAML

python

from skippr import Config, DataSourceRevolutBusiness, EnvRef, Pipeline

cfg = Config.discover()
revolut = cfg.data_source(
    "revolut",
    DataSourceRevolutBusiness(
        client_id="your-client-id",
        start_date="2024-01-01",
        issuer_domain="example.com",
        refresh_token=EnvRef("REVOLUT_REFRESH_TOKEN"),
        private_key_pem=EnvRef("REVOLUT_PRIVATE_KEY_PEM"),
    ),
)
cfg.pipeline(
    "treasury",
    Pipeline(data_source=revolut, data_sink=cfg.get_data_sink("lake")),
)
cfg.save()

bash

skipprd connect data-source revolut-business \
  --pipeline treasury \
  --name revolut \
  --client-id your-client-id \
  --start-date 2024-01-01 \
  --issuer-domain example.com \
  --refresh-token '${REVOLUT_REFRESH_TOKEN}' \
  --private-key-pem '${REVOLUT_PRIVATE_KEY_PEM}'

yaml

data_sources:
  revolut:
    RevolutBusiness:
      client_id: your-client-id
      start_date: "2024-01-01"
      issuer_domain: example.com
      refresh_token: ${REVOLUT_REFRESH_TOKEN}
      private_key_pem: ${REVOLUT_PRIVATE_KEY_PEM}

pipelines:
  treasury:
    data_source: data_sources.revolut
    data_sink: data_sinks.lake

With the CLI, connect the destination to the same pipeline with skipprd connect data-sink … --pipeline treasury.

Check the connection and run one pass:

bash

skipprd discover --pipeline treasury
skipprd sync --pipeline treasury --once

A successful run lands a row in revolut_sync_run_daily with status = ok and the streams it synced.

Revolut access tokens expire after a short time. Skipprd uses the refresh token and private key to get a fresh one whenever it needs it, so long-running pipelines keep working. You can instead set access_token to a token you obtained yourself, which is useful for a one-off run.

Options

Key	Type	Default	Description
`client_id`	string	Required	Client ID from your Revolut Business API certificate. Written to every row as `connection_id`.
`start_date`	string (`YYYY-MM-DD`)	Required	Start of the transaction window.
`lookback_days`	integer	`90`	Days subtracted from `start_date`. Each run reads transactions created from `start_date` minus this many days up to now.
`stream_profile`	string	`console_default`	Which streams to sync: `minimal`, `console_default`, or `full`. See What gets synced.
`streams`	list of strings	Not set	Sync exactly these streams instead of a profile: `accounts`, `transactions`, `transaction_legs`, `health`. Unknown names are ignored.
`min_query_interval_ms`	integer	`300`	Minimum milliseconds between transaction page requests.
`api_base`	string	`https://b2b.revolut.com/api/1.0`	Revolut Business API base URL. Use `https://sandbox-b2b.revolut.com/api/1.0` for the sandbox.
`refresh_token`	secret	Not set	OAuth refresh token, as `${ENV_VAR}`. Required unless you set `access_token`.
`private_key_pem`	secret	Not set	PEM-encoded RSA private key matching the certificate you uploaded, as `${ENV_VAR}`. Required with `refresh_token`.
`issuer_domain`	string	Not set	Domain of your registered redirect URI. Required with `refresh_token`. You can set the `REVOLUT_ISSUER_DOMAIN` environment variable instead.
`access_token`	secret	Not set	A Revolut access token, as `${ENV_VAR}`. When set, the refresh settings are ignored.
`privacy.mode`	string	`passthrough`	`passthrough`, `profile`, or `allowlist`. See Personal data.
`privacy.profile`	string	`passthrough`	`passthrough` or `upfoundry_safe`. Setting `upfoundry_safe` switches on profile mode even when `mode` is `passthrough`.
`privacy.drop_properties`	list of strings	`[]`	Extra field names to remove (case-insensitive) when the `upfoundry_safe` profile is active.
`privacy.keep_properties`	list of strings	`[]`	In `allowlist` mode, the only top-level fields kept on each row.
`privacy.hash_properties`	list of strings	`[]`	Accepted but not applied in this version.
`privacy.on_violation`	string	`drop`	`drop` or `deadletter`. Accepted but not applied in this version.

Environment variables

Key	Type	Default	Description
`REVOLUT_ISSUER_DOMAIN`	string	Not set	Used as the issuer domain when `issuer_domain` is not set in the config.

What gets synced

Stream	Tables	`minimal`	`console_default`	`full`
`accounts`	`revolut_account_snapshot`, `revolut_account_source_raw`	✓	✓	✓
`transactions`	`revolut_transaction_fact`, `revolut_transaction_source_raw`	✓	✓	✓
`transaction_legs`	`revolut_transaction_leg_fact`, `revolut_transaction_leg_source_raw`		✓	✓
`health`	`revolut_sync_run_daily`	✓	✓	✓

console_default and full currently select the same streams. revolut_sync_run_daily gets one row per run (run_date, status, streams_synced, elapsed_ms) whichever streams you pick.

Table	Contents	How it lands
`revolut_account_snapshot`	Account ID, currency, balance, state, created and updated times	One row per account per `run_date`. Each run replaces that day's rows.
`revolut_transaction_fact`	Transaction ID, type, state, currency, amount, created and completed times	Upserted by `transaction_id`.
`revolut_transaction_leg_fact`	Leg ID, transaction ID, account ID, currency, absolute amount, `direction` (`credit` or `debit`)	Upserted by `leg_id`.
`revolut_*_source_raw`	Record ID, stream, fetch time, and a SHA-256 hash of the record after privacy rules	Upserted by record ID and `ingest_run_date`. Use the hash to see when a record changed. The record body itself is not stored.

How runs build on each other. Every run reads all transactions created from start_date minus lookback_days up to now, then upserts them. A transaction that settles or is reverted after you first synced it is updated in place. Account balances are a fresh snapshot each run; filter to the latest run_date for current balances.

Deletes. Revolut doesn't delete transactions; a reversed or declined transaction shows up through its state.

Discover. skipprd discover reads only today's transactions and the minimal streams to learn column types.

Personal data

The curated tables never include account names, counterparties, references, descriptions, or card details. The default passthrough mode also removes top-level fields named exactly like personal data (name, reference, description, iban, email, …) before the record hash is computed.

Setting	Effect
`mode: passthrough` (default)	Removes top-level fields named exactly like personal data.
`profile: upfoundry_safe`	Also removes any field, at any depth, whose name contains a personal-data term such as `merchant`, `counterparty`, `beneficiary`, `card`, or `iban`, plus every field in `drop_properties`.
`mode: allowlist`	Keeps only the top-level fields in `keep_properties`. Always include each table's ID column and date column.

Troubleshooting

Symptom	Cause	Fix
`sink '…' does not support write policy MergeByKey …` or `… ReplacePartition …`	Your destination can't upsert or replace partitions.	Use Athena Iceberg or SkipprLake.
`access_token or refresh_token is required`	Neither credential reached Skipprd.	Export the variable named in `refresh_token` in the shell that runs `skipprd`, and check the name matches.
`private_key_pem is required to refresh access tokens`	`refresh_token` is set but the key is missing.	Export the PEM file contents into the variable named in `private_key_pem`.
`issuer_domain is required for JWT client assertion …`	No issuer domain is set.	Set `issuer_domain`, or export `REVOLUT_ISSUER_DOMAIN`.
`Revolut HTTP 401`	The refresh token was revoked or expired, or the key doesn't match the uploaded certificate.	Re-run Revolut's consent flow for a new refresh token, and check you're using the private key for that certificate.
`Revolut HTTP 401` against the sandbox	`api_base` points at production while your credentials are for the sandbox (or the other way round).	Set `api_base` to match the environment that issued the client ID.
`invalid start_date`	The date isn't `YYYY-MM-DD`.	Quote it in YAML: `start_date: "2024-01-01"`.
Runs get slower over time	Every run re-reads the whole window from `start_date` minus `lookback_days`.	Move `start_date` later once you have the history you need.

Revolut Business ​

Before you begin ​

Configure ​

Options ​

Environment variables ​

What gets synced ​

Personal data ​

Troubleshooting ​

Next steps ​

Revolut Business

Before you begin

Configure

Options

Environment variables

What gets synced

Personal data

Troubleshooting

Next steps