Case Study: Flo Health achieves 10x faster AI evaluations with SuperAnnotate

Flo Health cuts AI evaluation cycles by 10x with SuperAnnotate

Flo Health, the leading women's health platform, faced a challenge in operationalizing clinical expertise to safely evaluate its AI-powered assistant, AskFlo. The company needed to validate over 12,000 LLM outputs against strict clinical guidelines to achieve a greater than 90% accuracy threshold, but its manual processes created bottlenecks that stretched iteration cycles to weeks. To solve this, Flo turned to the vendor SuperAnnotate.

By implementing SuperAnnotate's expert-in-the-loop platform and integrating it with Databricks, Flo built a robust evaluation pipeline. This solution cut iteration cycles from 2-3 weeks down to 3-5 days, achieving a 10x faster evaluation throughput. The results included a validated accuracy of over 91% at scale, which allowed Flo to launch AskFlo with clinical-grade safety and confidence.

Flo Health