Implementing Guardrails using Semantic Similarity

August 1, 2023

Announcing Guardrail Alerts and Reporting

We are excited to announce new Guardrail Alerts and reporting.

With Reconify analytics, guardrails can easily be added to prompts and responses based on semantic similarity. Simply select a prompt or response, and create a guardrail alert.

Reconify analyzes incoming prompts and responses to check if they are semantically similar to the alert. If so, a notification will be sent in realtime via email or webhook to take action.

The new feature makes it even easier to monitor Generative AI applications and take action based on the analytics.

Check out the demo video below to see guardrail alerts in action.

About Reconify

Reconify enables developers and enterprises to analyze and optimize Generative AI to improve response effectiveness and customer satisfaction.