... I've learnt a lot since joining Damien Garros and the OpsMill team 🦦 last year... before then I may have sold hardware and software into technology firms but not necessarily the glue that holds it together... Some of you who have been around as long as I may remember the Sun Microsystems slogan "the network is the computer" which probably holds truer now than 20 years ago?! And yet a mere three years ago, suggesting your network source of truth needed branching, peer review, and CI might have got you some funny looks 🫨
This spring, well... not so much. Now that every IT and network leader wants to know how their team is going to “make the most of AI,” multiple long-established platforms have shipped those very features. Turns out clean, governed, version-controlled infrastructure is the prerequisite for almost everything AI is supposed to do.
Because if AI agents are coming for network automation, you can’t just hand them the keys to fragmented, stale underlying data.
Even if you have perfectly unified technical and business logic, you’ll still want every change agents make to be validated automatically against your schema, approved by a human, and easy to roll back if it’s wrong.
And when your small team is asked to start automatically spinning up new AI data centers (“You can do that, right?”) with hundreds of thousands of GPUs, liquid cooling, optical fabrics, and topologies that didn’t exist five years ago, you’re going to need a schema that can describe a fabric, a pod, a service, not just devices and ports.
The point is, you need a flexible schema, version control, and governance all in one platform, not tacked on one release at a time.
Read our full take on the blog "Why bolt on when you can build in" below ↓
https://lnkd.in/eVKD-GfE