Overview
The Alchemer 'Redact sensitive data' integration uses Google’s 'Sensitive Data Protection' capabilities to automatically detect and redact sensitive information from survey data. This integration allows organizations to protect personal and confidential data by identifying sensitive content and applying de-identification rules before data is stored, routed, or passed to downstream system.
Common uses for the Alchemer Redact sensitive data integration
- Detect and redact personally identifiable information (PII) from survey data
- Protect sensitive survey responses before routing or storage
- Apply consistent de-identification rules across workflows
- Prepare sanitized data for downstream integrations and analytics
- Reduce compliance risk when handling sensitive information
What can the Redact sensitive data integration do?
- Inspect survey data for sensitive content
- Redact or transform detected sensitive values
- Return sanitized outputs for use in survey
You will need
- An inspectConfig object defining what sensitive data to detect — or one of the built-in configuration levels described below
- A deifentifyConfig object defining how detected data should be redacted or transformed — or one of the built-in configuration levels described below
-
An Alchemer plan that includes integrations and the Integration Manager permission enabled.
- Contact us if you are unsure if your plan includes integrations.
Setup Redact sensitive data integration in workflow
Tools & Utilities | Redact sensitive data
You will need:
- A valid inspectConfig object, or one of the built-in configuration levels (Basic, Intermediate, or Advanced)
- A valid deifentifyConfig object, or one of the built-in configuration levels (Basic, Intermediate, or Advanced)
- Survey fields containing data to inspect and redact
Configure the action
- Open your survey in Survey builder.
- Select Add New: Action.
- In the Add Action modal, scroll to the Integrations section.
- Select Tools & Utilities.
- Select Tools & Utilities | Redact sensitive data.
- Tools | Select default configuration: Choose a configuration level from the dropdown—Basic, Intermediate, Advanced, or Custom. Each level pre-populates the Inspect Config and Deidentify Config fields below with a different set of sensitive info types and redaction approach. See Default configuration levels below for what each level detects and how it redacts matches. Select Custom to start from an empty configuration and build your own.
- Tools | Deidentify configuration: Review the
inspectConfiganddeidentifyConfigobjects in the respective textbox. These are pre-filled based on the configuration level you selected above. Edit either object directly if you need to adjust the info types detected, the likelihood threshold, or the redaction method.- Note: The Basic level pre-fills the deidentify configuration with an empty
redactConfigobject, which does not remove or mask matched values on its own. If you select Basic, you must still configure how matches are handled (for example, by populatingredactConfigor replacing it with a masking configuration) before matched values will actually be redacted.
- Note: The Basic level pre-fills the deidentify configuration with an empty
- Tools | Select question to redact: Select the field in this survey that you want to redact. Only one field can be redacted per action. Add additional actions as needed.
- Save the action.
Status codes
- 200: Successfully redacted sensitive data
- 400: The external integration returned an error
Default configuration levels
Each configuration level sets a starting point for what’s detected and how it’s redacted. You can edit the generated inspectConfig and deidentifyConfig objects after selecting a level, or start from scratch with Custom.
Note: Selecting a level only pre-fills the configuration objects—it doesn’t guarantee that matched values will be removed or masked. In particular, the Basic level generates an empty redactConfig placeholder in the deidentify configuration. Until you configure that object (or switch to a level like Intermediate or Advanced that includes a populated masking configuration), matched values will pass through unredacted even though they were successfully detected.
| Level | Info types detected | Likelihood threshold | Redaction method |
|---|---|---|---|
| Basic | Person name, email address, phone number, SSN, credit card number | Very likely (flags only high-confidence matches) | None set by default. The generated deidentifyConfig includes an empty redactConfig object as a placeholder, but matched values are not removed or masked until you configure the redaction behavior yourself. |
| Intermediate | Everything in Basic, plus street address, date of birth, IP address, passport number | Likely | Character masking — matched values are replaced with asterisks |
| Advanced | Everything in Intermediate, plus financial account number, medical record number, generic ID, auth token, password | Possible (flags more potential matches, which can increase false positives) | Character masking — matched values are replaced with asterisks |
| Custom | None by default — you define your own infoTypes |
Possible (editable) | None by default — you define your own transformation |
Testing & Support
Testing and Validation
How to test
- Submit a survey response that triggers the integration action.
- Confirm the expected field has been redacted in the results.
- Use metadata for verification and debugging
How to verify results
- Inspect returned metadata values for redacted content.
- Use sanitized outputs in routing, conditions, or merge codes.
Monitoring Integration Activity
Where to find logs
- Go to Results → Individual Responses.
- Select the Redact sensitive data integration step.
What logs display
- Inspection inputs and redacted outputs
Troubleshooting
Configuration issues
-
Invalid or improperly formatted
inspectConfigobject -
Invalid or improperly formatted
deidentifyConfigobject
Redaction issues
- Sensitive data types not included in the inspection configuration
- Unexpected redaction behavior due to de-identification rules
-
Values are detected but not redacted or masked—this is expected with the Basic configuration level unless the
redactConfigobject has been configured
FAQs
What permissions do I need?
Integration Manager permission in Alchemer.
Integration Manager permission in Alchemer.
When does the integration run?
When the workflow reaches the Redact sensitive data integration step.
When the workflow reaches the Redact sensitive data integration step.
Can I use multiple Redact sensitive data actions in one workflow?
Yes. Each action runs independently and can be chained as needed.
Yes. Each action runs independently and can be chained as needed.
Why isn’t my data being redacted?
Check that your
Check that your
inspectConfig includes the correct infoTypes and your deidentifyConfig defines valid transformation rules. If you selected the Basic configuration level, note that its default deidentifyConfig includes an empty redactConfig object—matched values are detected but not removed or masked until you configure that object yourself.
What if I need additional functionality?
Contact Alchemer Support for enhancement requests.
Contact Alchemer Support for enhancement requests.
