1000 AI agents that run a marathon together.
What does a multi-agent system look like at production scale?
And how many tokens does it cost?
Full video below
I’m going live on the Google Cloud Tech YouTube channel with my coworker Ayo to walk through an AI lab I built.
If you’re interested in building AI agents with multimodal data + memory, come join us!
Time: March 31, 9:00 AM PST
I just built a real-time conversational agent with bidi streaming.
It streams responses, updates memory in real time, and lets you interrupt anytime.
I wrote 2 blogs on how it works.
You don’t even need to read them. 🤣
Feed them into your AI assistant
and use it to build your own agent :)
What can you learn from #AIAgentBakeOff? Judges @ivnardini and @anniewangtech break it down:
1️⃣ Reasoning behind agentic patterns
2️⃣ Optimizing system design
3️⃣ Knowing when to put a humans in the loop
Watch the full episode → https://t.co/TEX4qvb7xQ
Did 3 talks in #Beijing this week.
#GoogleCloud#Summit → #BuildwithAI workshops.
The energy from the AI community in China is real, sharp questions, thoughtful builders, and genuine excitement to learn and build.
Grateful and motivated 🤍
How can I make my AI agent more deterministic?
Here’s my quick take 👇
If you’re building with ADK (Agent Development Kit), there are several ways to add more control:
1️⃣ Use workflow agents (Sequential, Parallel, or Loop) to structure execution flow
2️⃣ Create custom agents to define your own logic or conditions
3️⃣ Add callbacks to run custom logic before or after tool calls
4️⃣ Define fallbacks so your agent stays reliable when something goes wrong
#AIAgent #multiagent
How do you build AI agents that remember and learn from past conversations?
Watch this episode of The Agent Factory for a deep dive into building agents with memory, including the fundamentals of agent memory, Vertex AI Memory Bank, and more → https://t.co/MlRr6gUKmp
Excited to share my first time recording on the Google Cloud podcast, The Agent Factory 🎙️
👉 https://t.co/ExecuYOZnQ
Together with @ivnardini and Kimberly Milam from Cloud AI Research, we dive into Agent Memory & Vertex AI Memory Bank:
- Agent short-term vs. long-term agent memory
- How memory enables personalization & collaboration (beyond static RAG)
- How memory & retrieval work in real projects
Would love to hear your thoughts : )
Hey, happy Friday :)
Can you believe you can design, build, test, and deploy an AI agent, all with Gemini CLI?
Just: 💡 Idea → 🤖 Agent → 🚀 Deployment
I created a hands-on workshop to walk you through the full process https://t.co/kjiKTHKls1
See you there at 10:15
How do you build a multimodal AI agent?
I just wrote a blog showing how to use Genkit to build a voice-input accessibility agent (speak → understand → act → respond).
The blog here: https://t.co/aTMQQshMDh
#AI#Agents#Accessibility#Genkit#Multimodal
On August 22, I’ll be speaking at a Google Cloud Lab Workshop, where we’ll walk through how to turn an idea into a production-ready AI agent, covering the full end-to-end SDLC with agents.
✨ Think of it as using magic… to create magic.
RSVP: https://t.co/kjiKTHKThz