Case Study: Wellthy achieves secure AI testing and 50% fewer flagged actions with Tonic.ai

Wellthy cuts flagged actions 50% with Tonic.ai

Wellthy, a fast-growing digital healthcare company, faced a challenge in developing new generative AI features due to limited access to realistic yet secure test data. Their unstructured data, such as free-text messages containing sensitive PII and PHI, was particularly difficult to anonymize effectively without losing realism, which hindered their testing and development processes. They turned to vendor Tonic.ai and its Tonic Textual product for a solution.

Tonic Textual provided Wellthy with a way to de-identify and synthesize their unstructured data, creating realistic but safe messages for testing. This solution allowed Wellthy to confidently develop and launch new AI features. The results were significant, including a genAI feature that cut down on unnecessary workflow action items by 50%, eliminating tens of thousands of redundant tasks. The use of Tonic.ai's realistic test data also led to fewer bugs reaching production and gave developers better context for the product they were building.

Wellthy