Quality Engineering for GenAI

AI-based systems today are used for all sorts of applications. Already we have learned that the quality engineering activities for AI-based systems differ from what we are used to for traditional systems. In this part of the TMAP website you will find how quality engineering and testing activities can be performed to get the right quality at the right moment for the AI-based system.

Areas of interest for testing AI.

Quality Engineering for AI - Data / Input

Quality Engineering for AI - Answer Generation

[Under construction]

Wiki:
Machine Intelligence quality characteristics

or download the original paper.

	Data / Input	• Test data consistency on ingestion • Intent Recognition Testing • Entity Extraction Testing.
	Answer Generation	• Confidence Level Testing • Multi-Turn Conversation Testing • Guardrail testing / unhappy flow testing • Fallback Scenario Testing • Edge Case Testing.
	System	• Backend performance / stress tests for latency • User acceptance tests • Testing for monitoring and logging • User feedback testing.