How to build an accuracy pipeline for your AI app

4 459
11
Следующее
Популярные
93 дня – 13 7487:39
Connecting ADK Agents to MCP Servers
Опубликовано 2 февраля 2026, 17:00
Evaluating agents with ADK code lab → goo.gle/3NRVhSB
Evaluating single LLM outputs With Vertex AI evaluation code lab → goo.gle/4jYfYZ0

Large language models (LLMs) are great but sometimes they go rogue. How can developers gain real confidence in their AI systems while in production? Join Aja and Jason as they demonstrate how to implement an 'accuracy pipeline' using LLMs as your ultimate grading rubric, treating each evaluation prompt like a shiny new unit test.

Chapters:
0:00 - Intro
0:48 - What is hallucination?
2:35 - Testing AI answers for accuracy
5:43 - Offline evaluation
8:53 - Summary

More resources:
Agent evaluation in Vertex AI Gen AI evaluation service → goo.gle/3M0l3Dw

Watch more Real Terms for AI → goo.gle/AIwordsExplained
🔔 Subscribe to Google Cloud Tech → goo.gle/GoogleCloudTech

#GoogleCloud #AIInfrastructure

Speaker: Aja Hammerly, Jason Davenport
Products Mentioned: AI Infrastructure
автотехномузыкадетское