Security | PII & PHI anonymization

When it comes to qualitative feedback ingestion, anonymizing PII and sensitive information is a key step to keep your data pipeline secure and privacy compliant.

• Personal numbers or codes such as a United States Social Security Number, driver's license. Or any document from other countries.
• ⁠A person's full name or first name.
• ⁠A person's phone number.
• ⁠A person's email address.
• A person's address, location, or city.
• ⁠Any credit card, bank account information, or passwords.

A full list of removed PII & PHI can be found at the bottom of this article.

By design, the Birdie solution does not rely on PII or sensitive information to work, therefore, all native connectors in Birdie are created to import only anonymized structured columns. It means, instead of importing user names and emails from your support ticket platform, Birdie connectors will only import user_ids to match the records.

More than structured columns, there may be PII or sensitive information within qualitative fields. In cases like that, there are two different directions to follow:

Integration flow with PII anonymization handled by Birdie

• In this scenario, Birdie will handle PII anonymization within the initial step of data pipeline in GCP region: us-central-1
  
  

• Our anonymization process employs a hybrid system combining rigid pattern detection with a AI-based entity recognition model. This means that we have two layers of pattern matching for specific IDs to successfully remove most PIIs.
  
  

• Any sensitive pattern such as: names, emails, addresses, card numbers, document ids will be [REDACTED] to indicate the change in the original content

• PII anonymization is activated for all clients, ensuring that all incoming open-text fields (ticket comments, survey responses, call transcripts) will get checked and processes before Birdie stores any raw data in our servers

• This option can be used with any ingestion method (native connectors, S3 bucket, Birdie Rest API or CSV uploads).However, there is the need to send open PII to be anonymized

Note: For clients in Brazil, in compliance with the Lei Geral de Proteção de Dados (LGPD), this ingestion step runs in GCP region: southamerica-east-1 prior to final persistence at the global index in GCP region: us-central-1, Iowa. For clients outside Brazil, the whole process will remain running at our main infrastructure in the United States, GCP region: us-central-1, Iowa.

Integration flow with PII anonymization handled by your company

• In this scenario, your company handles PII anonymization within your local infrastructure, running PII redaction scripts to replace names, emails, addresses, document IDs and whatever is considered sensitive by regulations in your industry/business

• To help running that task internally, we offer this anonymizer project, available in the Google Cloud Marketplace

• Once anonymized, data integration can be done via S3 bucket (parquet file) or Birdie Rest API

• By using this approach, your company make sure Birdie never receives any PII or sensitive information

• Names (including doctor names)
• NINs (National Identification Numbers)
• Driver's License Numbers
• Passport Numbers
• Patient Account Numbers
• Birth Dates
• Admission/Discharge Dates
• Vehicle Identifiers
• Device Serial Numbers
• URLs and IP Addresses
• Biometric Identifiers
• Voter Card IDs
• Taxpayer Identification Numbers (TINs)

• Phone Numbers (including fax)
• Email Addresses
• Physical Addresses
• ZIP Codes
• Certificate/License Numbers

• Credit Card Numbers
• Bank Account Numbers
• Insurance Policy Numbers
• Payment Details

• Medical Record Numbers (MRN)
• Insurance IDs and Group Numbers
• Medicare/Medicaid Numbers
• DEA Numbers
• NPI (National Provider Identifier)
• Health Plan Beneficiary Numbers
• Device Identifiers and Serial Numbers
• Laboratory Numbers

• Diagnosis Codes (ICD-10)
• Procedure Codes (CPT)
• Medication NDC Codes
• Test Results
• Treatment Dates
• Appointment Details