Multi-Agent AI, Open Protocols & Startup Acceleration with Saurabh Tiwary | AI Basics

24 snips

Jun 26, 2025

Saurabh Tiwary, VP & GM of Cloud AI at Google, shares insights on the cutting-edge world of multi-agent systems. He discusses how AI agents are evolving from simple task-doers to complex decision-makers. The conversation highlights Google’s open frameworks, including the Agent Development Kit and the Agent-to-Agent Protocol, which enhance collaboration across platforms. Saurabh also emphasizes the significance of multimodal inputs and the need for human oversight to ensure responsible AI usage, promising exciting advancements in startup acceleration.

Ask episode

AI Snips

Chapters

Transcript

Episode notes

INSIGHT

Evolution of AI Agents

AI agents are evolving to handle indeterminate, multi-step tasks autonomously.
They analyze outputs and decide subsequent actions without explicit programming for each step.

ADVICE

Build Trust with High Accuracy

Deploy agents only when their task accuracy exceeds 80-90% to build trust.
Gradually increase agent task complexity as confidence in their outputs grows.

ANECDOTE

AI Helps Parse Founder Updates

Jason's team uses AI to extract key metrics from founder updates, saving 15-30 minutes per review.
The AI identifies missing data but human responses remain necessary for now.

Get the Snipd Podcast app to discover more snips from this episode

Get the app