canonical: https://jentic.com/apis/nexmo.com/nexmo-redact-api # Nexmo Redact API The Nexmo Redact API gives privacy and compliance teams a controlled way to scrub the personally identifiable content of a specific transactional record from Vonage's logs. A single POST endpoint accepts a transaction id and product type and removes the message body, recipient number, and other PII from Vonage's retained records, helping organisations meet GDPR right-to-erasure and similar privacy obligations on demand. ## For AI agents Redact the personally identifiable content of a specific Vonage transactional record (SMS, voice call, verification, message) from Vonage's logs. ## Scope Does not send messages, manage user consent records, or audit access logs - use for redacting specific Vonage transactional records only. ## Capabilities - Redact a specific SMS transaction from Vonage's retained message logs - Redact a specific voice call record from Vonage's retained call logs - Redact a specific verification request from Vonage's retained Verify logs - Redact a specific Messages API record across SMS, MMS, WhatsApp, and Viber - Honour user GDPR right-to-erasure requests programmatically against Vonage's logs ## Use cases ### GDPR right-to-erasure on transactional logs When a customer submits a right-to-erasure request, the compliance team identifies the affected SMS, voice, or verification transactions in their own systems, then calls POST /transaction with each id and product to scrub PII from Vonage's retained logs. Closing the loop on Vonage's side means the organisation can attest that no copy of the message body or number remains in any system under its control. Example prompt: POST /transaction with id=MESSAGE_ID and product=sms for each message tied to the data subject ### Voice call PII scrubbing Customer support and trust teams that record voice calls or store recipient numbers in Vonage's retained logs can call the Redact API after a complaint or breach to remove the PII associated with a specific call. This is critical when a recording has been flagged as containing sensitive information that should not have been captured. Example prompt: POST /transaction with id=CALL_UUID and product=voice for the affected call ### Verify and Messages API cleanup Organisations using Vonage's Verify or Messages APIs can redact a specific verification or rich-message record by transaction id. Because the endpoint requires the original transaction id, redactions are scoped and auditable - you cannot accidentally erase records you do not specifically reference. Example prompt: POST /transaction with id=REQUEST_ID, product=verify, and type=verify-request to scrub the verification record ### AI agent compliance automation via Jentic An AI agent that handles privacy ticket workflows can call the Redact API through Jentic each time a right-to-erasure request is approved. The agent reads transaction ids from its own ticket store, searches Jentic for the redact operation, and executes one POST /transaction per record, all with credentials supplied by your Jentic One instance. Example prompt: Search Jentic for 'redact a vonage transaction', load POST /transaction, and execute with id and product for each ticket-linked record ## Key endpoints | Method | Path | Description | | --- | --- | --- | | POST | /transaction | Redact a specific SMS, voice, verify, or messages record by id | ## Key resources - **Transaction redaction** — POST /transaction redacts PII from a specific Vonage transactional record by id and product. ## Why Jentic - **Setup:** Wiring the Redact API by hand means encoding your api_key and api_secret as an HTTP basic header on every request and posting redaction requests to the api.nexmo.com/v1/redact host. Through Jentic you install once, import the Redact API from the API Directory, store the credentials once, and your agent calls it. - **Permission scoping:** The Redact API exposes a single endpoint and carries the transaction id and product in the request body, so scoping is by operation rather than by resource: limit the agent to the operation it needs, which is redacting a transaction. That single redaction call is the only thing it can run. - **Credential handling:** Your Nexmo api_key and api_secret are stored once, encrypted, by your own Jentic One instance and encoded as HTTP basic at execution time. They never enter the agent's prompt, logs, or context. - **Discovery method:** Agents search Jentic by intent such as 'redact a Vonage transaction', and Jentic returns the POST /transaction operation with its input schema so the agent calls the right endpoint without browsing the reference docs. ## Related APIs - **Twilio Messaging** — Twilio offers per-record deletion on its messaging resources as the closest equivalent compliance hook. - **Nexmo SMS API** — SMS messages sent via the SMS API are the records the Redact API later scrubs. - **Nexmo Verify API** — Verify transactions are one of the four product types the Redact API can scrub. ## FAQ ### What authentication does the Nexmo Redact API use? The Redact API uses HTTP Basic authentication with your Nexmo apiKey as the username and apiSecret as the password - different from most other Vonage APIs that take credentials as query parameters. Through Jentic those credentials are encoded automatically and the agent only ever sees a scoped token. ### Can I redact more than one record in a single request? No - POST /transaction redacts exactly one record per call, identified by the id and product fields. For bulk erasure pipelines, loop through the transaction ids in your ticketing system and post one redaction per record. The request is intentionally narrow to keep the audit trail tight. ### What are the rate limits for the Redact API? The OpenAPI spec does not encode a numeric rate limit. The endpoint is intended for compliance workflows, not high-throughput automation, so a sensible client serialises requests and reconciles failures rather than running unbounded concurrency. ### How do I redact a transaction through Jentic? Run pip install jentic, search for 'redact a vonage transaction', and Jentic returns POST /transaction. Execute with id=TRANSACTION_ID and product (sms, voice, verify, or messages) and Jentic returns Vonage's confirmation. ### What products can the Redact API target? The product field accepts sms, voice, verify, and messages, covering every PII-bearing log Vonage retains for transactional traffic. The optional type field further narrows verify requests into verify-request, verify-check, and similar subtypes. ### Is the Nexmo Redact API free to call? Vonage does not charge per redaction request - the API exists to support compliance obligations. The cost of the original SMS, voice, or verification transaction is unaffected by a later redaction. ### Can I limit what my agent is allowed to do with the Nexmo Redact API? Yes. The Redact API exposes a single operation, POST /transaction, and because you run Jentic One yourself, your own rules decide whether the agent may call it and which credentials it uses. You can restrict the agent to just that one redaction operation, so the only action it can perform is scrubbing a specific Vonage transactional record by id and product. Your Nexmo api_key and api_secret stay stored and encrypted on your own instance and are encoded as HTTP basic only at execution time, never entering the agent's prompt or logs.