introduction

Author: ashpreetbedi

evals/introduction.mdx

@@ -4,12 +4,12 @@ title: Introduction
 
 ## What are Evals?
 
-**Evals** are like tests for your Agents. Use them judiciously to evaluate the performance of your Agents and improve them over time.
+**Evals** are unit & integration tests for your Agents. Use them judiciously to evaluate and measure the performance of your Agents and improve it over time.
 
 We typically evaludate Agents on 3 dimensions:
 
-- **Accuracy:** How accurate is the Agent's response?
-- **Performance:** How fast does the Agent produce the output and what is the memory footprint?
+- **Accuracy:** How complete/correct/accurate is the Agent's response (LLM-as-a-judge)
+- **Performance:** How fast does the Agent respond and what's the memory footprint?
 - **Reliability:** Does the Agent make the expected tool calls?
 
 ### Accuracy