IBM just released a 1-hour course on building agentic knowledge graphs from scratch: • 00:00 - Introduction to knowledge graphs • 05:35 - Building your first agentic graph • 19:59 - Agentic memory powered by graphs • 30:39 - Graphs for multi-agent orchestration This 1-hour watch will replace 10... paid courses on agentic engineering. Watch it today, then learn how to become a knowledge graph engineer in the article below.show more

这种讲解风格 好像王自如也做过，还讲解过怎么制作 应该是人单独录一版 书写一版，再剪辑制作 不然从正面看 人的书写，观众视角来看，字应该是反过来的

that 30:39 multi-agent orchestration section sounds like the standout here

Basiclly Graphs are made for Multi-agents orchestration, and this part is a pure gold

his writing is really easy to read i like this approach, thanks for sharing

Yeah, basically knowledge graphs explained simply.

one hour covers the whole graph stack

Yeah, true - if you watch this, you’re basically at master degree level in graph engineering. Just apply it to your agent stack and you’re gucci.

Thanks for sharing, though pretty sure this was from 4 years ago?

thanks for the courses, I added them to my bookmarks

You are welcome Evgenia. Whats your favorite btw ?

wow man, that’s a pure gold material on graph engineering!

Yeah, IBM has been shipping real alpha for years, and this watch is one of them.

This is what I needed

All you need to master Graphs is to spend 1 hour on this video!

Graphs earn their keep at reconciliation, not recall. Storing what happened is easy. Deciding which of two conflicting facts survives is the part that needs structure, and where most memory layers quietly do nothing.

In order to deeply understand multi-agent orchestration graphs, it’s so important to build a strong foundation with the core agent loops and “loops” like /goal and /autoresearch . @christinetyip and I review these fundamentals before discussing subagent graphs in our latest episode.

@0xCodez 1 hour isn't bad for all that. Do they go deep into agentic memory? That's what I'm most curious about.

ibm really said let us cook one more time 1 hour to replace 10 courses is either genius or the other 10 were just padding lol

@Whimsicali you seen this?

Just release? can you provide the link?

Dude, I went to watch it on youtube to realize that video is 4 years old.

@IBMwatsonx 这种已经不在牌桌上的教别人用Agentic graphs，有时间把自己的Bob产品打磨好。

Damm, I guess I was too early in 2023 talking about it…

Multi-agent systems eventually become a state-management problem. Chunks can store information, but graphs can represent who knows what, what changed, and why a decision was made.

Thanks for sharing.

An hour to replace 10 courses, bookmarking this one for the weekend 📚

Organizations adopting neural networks typically succeed when they integrate it into broader strategic goals in emerging economies.

Knowledge graphs are becoming a powerful foundation for building smarter ai systems. Combining structured knowledge with agent workflows can unlock more reliable reasoning, memory, and multi agent collaboration.

Graphs add value when they improve reasoning, not just storage. Connecting entities, relationships, and provenance gives agents a stronger foundation for retrieval and decision-making.

knowledge graph engineer wasnt even on my radar last year now its looking like the move fr.

Yea...lets listen to the Watson guys about how to build shit. Dont get me wrong appreciate the thought behind it but you have actual industry leaders dropping thought pieces and instructional content every other day lmfao

I confirm this is real α

Yeah I did this already, but I’ll see if there are any nuggets I missed.

one hour can teach the pattern - production graphs still need schema migrations, provenance and failure recovery

this replaces a stack of paid courses, save it now

bookmarking this to pretend i will watch it this weekend

martin keen at a whiteboard is the most trusted format in this whole niche

This is how a typical agent orchestrator looks like managing your tools, memory and agents.

Knowledge graphs are having a moment because everyone hit the same wall with RAG. Chunks give you recall, graphs give you provenance. The moment you need to say where a claim came from, the graph is the only one that answers.