Tokenization

Health and Pharmaceutical Use Cases

Anonymised Patient Data for Medical Research

A healthcare organisation collects Personally Identifiable Information (PII) as part of patient records, such as names, dates of birth, and diagnoses. A unique token is generated for each patient which allows the provider to track patient records across systems without revealing the actual identity of the patient. Researchers can access large volumes of health data to conduct studies on without compromising patient privacy.

Clinical Decision Assistance

A hospital uses a private Large Language Model (LLM) to assist physicians with treatment guidelines while maintaining strict patient data confidentiality. The LLM can generate suggestions based on the latest medical research and aggregated data, without directly accessing identifiable patient data.

Financial Services Use Cases

Anonymised Customer Data for Fraud Detection

Financial organisations collect sensitive customer data, such as names, account numbers, and transaction history, in order to detect for fraud. To protect this information, key parts of the data are tokenised which means it is replaced with random tokens that represent the original data without exposing it. The anonymised encrypted dataset can now be used for fraud detection without exposing the original sensitive data.

Personalised Financial Advice

A financial institution can utilise a private LLM to assist customer service representatives by generating responses to complex customer queries, such as questions about loan options without exposing customer data to external servers. The private LLM is trained on customer interaction data, policy documents, financial regulations etc all within a secure, on-premises, or private cloud environment.