Features & Roadmap

• Diverse transformation techniques: Generalisation, masking, hashing, encryption, pseudonymisation, suppression, and more
• K-anonymity testing: Test against IM8 WOG Access Modes A and B with privacy risk scoring
• Utility analysis: Assess information loss before committing to transformations
• Templates: Save, reuse, and share anonymisation configurations across your team for repeatable jobs
• Data tagging: Tag columns as direct identifiers, indirect identifiers, or sensitive attributes with guided sensitivity classification

• Auto-detection of 20+ entity types including names, NRICs, phone numbers, addresses, emails, dates of birth, bank accounts, and more
• Multiple input formats : Supports CSV, PDF, and DOCX files
• Custom entities: Define your own detection rules using inclusion lists, regex patterns, or privately-hosted LLMs for domain-specific terms
• Confidence tuning: Adjust detection sensitivity per entity type to balance recall and precision for your dataset
• Flexible anonymisation techniques: Redact, Replace with unique consistent tokens, Alias, Encrypt, Pseudonymise and more

Generate mock data that mimics the structure and format of real datasets.

• Singapore-specific fields: Names, NRICs, addresses, phone numbers, and other locally relevant formats
• Custom fields: Define your own data categories for domain-specific generation
• Field categories: Person, location, business, healthcare, and custom

Enable multi-party collaboration without exposing raw data.

• Secrets Manager: Create, store, share, and audit encryption keys and salts in a secure, centralised location
• Consistent encryption across datasets: Multiple data owners anonymise independently using a shared secret, producing data that can be linked on a common encrypted identifier
• Decryption: Authorised users with access to the secret can decrypt data when needed through the Web UI or API
• Audit trail: Full visibility of who used which secret and when

The same anonymisation capabilities available through the Web UI, designed for system-to-system integration.

• Three security levels: L2 (personalised token for integrated platforms), L3 (system token for GCC), L4 (signature-based for internet)
• Batch processing: Handle large volumes programmatically
• Advanced parameters: Custom recognisers, fine-grained confidence tuning, and allow lists via code

• Improve performance of free-text anonymisation through novel LLM-based approaches
• Extend modalities of access to serve a wider range of users, including non-government
• Incorporation of other privacy technologies, e.g., Homomorphic Encryption and Searchable Symmetric Encryption to process and search data even when encrypted