Multi-agent network ops gets much more interesting when agents can actually act on the network.
One of our speakers, Roger Perkin shares a quick preview π
On Oct 10, weβll build specialist agents with validation + human approval before changes.
Checkout the workshop here and get exclusive speaker discount ticket price at 35% - eventbrite.co.uk/e/multi-ageβ¦
Most "add AI to the platform" efforts fail the same way: hallucinations, runaway cost, security incidents. Then the verdict: agents aren't for us.
Kaspar Von GrΓΌnberg says that verdict is usually wrong. The model rarely breaks. The platform underneath does.
On Sept 30 he's showing the three-layer architecture that fixes it (tooling, paths, agent infrastructure) and what to build first if you already run an IDP.
Full disclosure: I help run these Packt workshops. This one's different because it starts with your platform, not with picking a model.
Since you are checking this out, I have a gift for you. Get the ticket of the workshop at flat 30% discount - eventbrite.co.uk/e/architectβ¦
Working in modern infrastructure often means moving between #AWS, #Azure, containers, #Terraform, #DevOps, and observability.
The Ultimate #Linux & Cloud Infrastructure Bundle brings practical books on all these areas together.
Take a look
humblebundle.com/books/ultimβ¦
π€ Made with AI
24 books. One bundle. The skills that run modern IT.
The Ultimate Linux & Cloud Infrastructure Bundle brings together practical Packt books covering five key areas:
π§ Linux and Windows administration
βοΈ Cloud infrastructure
π Networking
βοΈ DevOps and platform engineering
π Monitoring and observability
From managing servers and cloud platforms to automating infrastructure and troubleshooting production systems, the bundle covers the skills infrastructure professionals use every day.
Humble Tech Book Bundle: The Ultimate Linux & Cloud Infrastructure Bundle by Packt
AI models get the attention.
GPU infrastructure gets the bill.
On Sept 17, weβre going hands-on with what it actually takes to run AI workloads efficiently on Kubernetes:
β GPU scheduling & sharing
β Autoscaling
β Observability
β vLLM inference
β Keeping expensive GPUs from sitting idle
If you work in DevOps, Kubernetes, Platform Engineering or MLOps, this is the infrastructure side of AI worth learning.
ποΈEngineering GPU Infrastructure for AI Workloads Β· Luma
Genuine question for the network engineers here.
I book technical workshops for a living, so I spend most of my week listening to netops people argue about AI agents. Same complaint keeps coming up:
one agent with 12 tools doesn't investigate an incident. It guesses and it guesses with write access.
The fix everyone independently lands on has the same shape. A coordinator that owns the investigation, narrow specialists under it (one reads state, one proposes config, one validates), and a human gate before anything gets pushed.
Is that matching what you're actually running, or is it still mostly demo-ware outside of lab environments?
Punchy3 hrs 45 mins. Zero drop-off.
Senior directors. R&D engineers. All mid-conversation β MCP servers, Docker builds, Linux CLI β thinking alongside the session, not just watching it.
Some rooms you host. Some rooms host themselves.
This was the second kind. π
Excited to share that Eduard Dilharu is joining our upcoming networking workshop as a panelist.
For 3+ years, heβs been building AI agents for network engineers β from OSPF troubleshooting systems to multi-agent architectures powered by fine-tuned reasoning models.
A reader of an OpenShift book wrote:
βThe physical copy triggers a part of your brain differentlyβ¦ the platform just makes more sense.β
That stayed with me.
Maybe books do what screens often canβt:
No tabs.
No notifications.
No endless scrolling.
Deepak K retweeted
Thank you @PacktPub for this awesome book by Madhu Ramanujan "Salesforce CPQ Implementation Handbook"
It's always good to keep up with #Salesforce products with an eye focused on the more pragmatic part...less theory more action!
πamzn.to/3GT1HJJ
#neverstoplearning
Introducing the GKE Autopilot Cluster
{ by @iASHeeesh } from @hashnode. GKE Autopilot is nodeless managed Kubernetes service by @GoogleCloudTech.
#kubernetes #googlecloud #springboot ashishtechmill.com/introduciβ¦
Thanks @gnomeslair . Glad that you found the book helpful π
Yes, definitely a very useful book. amazon.com/Dynamic-Story-Scrβ¦
Wow!!! Thank you so much for the review. Great timing as the book is officially published today. Grab your copy now.π
amzn.to/3aFffsZ
Deepak K retweeted
Celebrate the joy of 2 hockey teams in Olympic S-F, remember to send a humble thank-you note to Odisha and its CM @Naveen_Odisha ....For 6 years now, the not-so-rich state has sponsored & nursed both men and women's hockey teams. If only more states were to spend on sport..











