Your AI agent still has no memory? Fix it with these 3 patterns
Annie’s codelab → https://goo.gle/4wA3kpd
Annie's code repo → https://goo.gle/3R39eyR
If your AI agent feels "dumb" because it can't remember a conversation from two minutes ago, you’re dealing with the Goldfish Problem. In part 1 of this series (https://www.youtube.com/watch?v=b9Dx7uGxwsg) Annie and Martin broke down the three basic memory patterns: session state, multi-agent state, and persistence. In this video they walk through three advanced patterns: Callbacks, Custom Tools, and Multimodal Memory. Use them to build personalized agents that impress your users!
Chapters:
0:00 - Intro
1:42 - Callbacks
3:29 - Custom Tools
5:30 - Multimodal Memory
6:42 - Recap
Watch more Serverless Expeditions → https://goo.gle/ServerlessExpeditions
🔔 Subscribe to Google Cloud Tech → https://goo.gle/GoogleCloudTech
#Serverless #GoogleCloud
Speakers: Martin Omander, Annie Wang
Products Mentioned: Agent Development Kit
Watch on YouTube ↗
(saves to browser)
Sign in to unlock AI tutor explanation · ⚡30
More on: Agent Foundations
View skill →Related AI Lessons
⚡
⚡
⚡
⚡
10 Real-World AI Agent Projects
Medium · LLM
Actually, vibe coding didn't kill testing — agentic engineering did
Dev.to · Muggle AI
Gemini 3.1 Flash Lite vs DeepSeek V4 Flash: Budget API Showdown for High-Volume Agent Loops (2026)
Dev.to AI
WebMCP Reality Check: Where the Spec Actually Stands
Dev.to AI
Chapters (5)
Intro
1:42
Callbacks
3:29
Custom Tools
5:30
Multimodal Memory
6:42
Recap
🎓
Tutor Explanation
DeepCamp AI