AI evaluation
Concept
The practice of systematically testing and measuring the outputs and decisions of AI models and agents, for example correctness, groundedness or orchestration quality, before and during production use.
| Type | Concept |
|---|
Articles about AI evaluation (2)
- Как тестировать AI-агентов и расширения Copilot воспроизводимым способом
· Mirko Peters
Как тестировать AI-агентов воспроизводимым способом: составить набор тест-кейсов, выбрать граудеры, оценивать вызовы инструментов и отслеживать регрессии, с примером в Copilot Stud - How to test AI agents and Copilot extensions in a repeatable way
· Mirko Peters
How to test AI agents in a repeatable way: build a test set, pick graders, score tool calls and track regressions, with a Copilot Studio example for QA leads.
Community wiki
Short info above, community knowledge below: add practical tips, examples or corrections. New contributions are reviewed first; trusted members publish directly.
Experts and companies for this topic
- Mirko Peters · Microsoft MVP
Sources
- Feed for this topic (RSS): entity/ai-evaluation/feed.xml
- Articles on this topic: 2
Maintained by Mirko Peters (Microsoft MVP) · Last updated: