Google just released the best 1-hour course on Graph Engineering: from single agent to a full 24/7…
145,312 次观看
15 条评论
评论来自 X (Twitter) 上的原帖

@DanielMiessler
Stolen Authority post. The person you linked to didn’t endorse this content. It’s your content appended to look as if they did. Shame.
7 个赞

@FromWageToOwner
Who is she?
1 个赞

@KevinOnFintech
Could you share the link to the Google course?
1 个赞

@kaddisdeployed
A free course covering all of this is honestly nice
1 个赞

@aukinhluan
Automation making use of AI language models?
1 个赞

@LivesMatters5
Graphs is just UML.

@MichaelQuanLoh
that's actually insane. This will definitely take my ai workflow to the next level

@0xRicker
really useful stuf bro

@0xNeoNat
Spot on! Loop-based agents are great for basic linear tasks, but graph engineering is essential once you need control, predictability, and parallel state handling

@omafunga
这波真的可以 图工程讲得挺清楚 从单agent到24/7系统 免费教程里算良心了 收藏了有空看看

@websterweby
saving this, the self-improving graphs part is what got me

@magsimich
graph engineering is the meta

@dstevens33700
The self-improving graphs part at 52:21 is the one I'm skeptical about. In practice they drift fast without a human checking what the graph actually changed before it goes live.

@0xlexlab
Carl Hewitt shipped the Actor Model in 1973: independent nodes, message passing, no shared state. LangGraph and every 2025 framework is Actor reheated.

@RemoteBrewser
The self-improving graphs part is the real unlock. Curious if they cover failure recovery or just optimization loops?
