Google just released a free 2-hour course on Graph & Loop engineering: 1 prompt → 100 agents → loops → graphs, from 0% to 100%: 0% → 0:35 - graph engineering from scratch 30% → 31:17 - build your first agent graph 45% → 43:40 - run hundreds of... agents in parallel 75% → 1:04:58 - loop engineering: route, check, repeat 100% → 1:25:09 - self-improving graphs that work while you sleep taught by the people who run this in production Prompts → Agents → Loops → Graphs most people keep stuffing instructions into one agent, and the real upgrade is one layer up same model, same tokens, and the only thing that changes is the shape you run it in watch it today, then save the full graph engineering guide belowshow more

wow I love this topic

@hanakoxbt honestly never thought I'd need to deep dive into graph engineering, but now I see how running parallel agents can elevate workflows. Might give this course a shot!

Running hundreds of agents in parallel is where the graph gets interesting. The hard part isn't spawning them, it's the fan-in: partial failures, retries, and keeping state consistent when one branch stalls. How are you handling that merge step?

Interesting concept for scaling AI agent capabilities.

100 agents running in parallel, wild scale

I have watched the complete video course and created a skill for implementing the learnings immediately in all my agentic AI workflows. If someone wants that I'll happily share. Just DM me. "Skill"

100 agents in parallel is the bit i’d watch first

One agent works. A loop makes it compound.

刚好最近在做多智能体架构，这个点说到点子上了，我这就去找课程来学学怎么搭分层结构。

Just two hours of your time, and you become better.

Beautif course

The real jump isn’t a better prompt. It’s going from 1 agent stuffed with instructions to a graph that fans out, checks itself, and keeps running while you sleep.Same model. Different shape.