Ok, I’m struggling to sit here at my desk—third coffee in hand—after just crawling back from DefCon (the hardware hacking village? Mind blown). It got me thinking about something near and dear to me: scaling NOCs for fast-growing companies. For those of you who don’t know me, I’ve been around this game since the year 1993, when I began as a lowly ol’ network admin after your momma just finished churning butter while dealing with the dandy ol’ PSTN muxes and voice/data lines… yes, well before the Slammer worm was a dark cloud knocking on the door of corporate networks. Now I run PJ Networks, and all we do here is cybersecurity, NOC, managed firewalls, servers, routers — whatever. This time, I’m exploring some of the trends and best practices I’ve been using for scaling a NOC. This is not just theory; it’s based on my dirty, thrilling experience — like when I worked with three banks on zero-trust architecture rework, except for the other time when I almost tore my hair out by the roots tuning toolchains for a growing fintech startup.
Here’s the thing. The larger your company, the more magical your network does not become easier to manage. You’re going from a handful of devices up to dozens, even hundreds of endpoints; data is flying into your systems as if it were rush hour in Mumbai. And if that NOC isn’t scalable, you’re going to be lost in an ocean of alerts, chasing false positives or worse – not even seeing the true threats.
I get it—automation is sexy. But you can’t replicate knowledgeable staff entirely. Consider it as the difference between driving a high-performance sports car such as a Porsche. Of course, you could throw in cruise control, assisted steering and the like, but someone who knows how to drive remains an important part of the equation (for intricate turns and unexpected obstacles).
Recruiting is one of the ways to do this (you should always be recruiting; note that two successful hires a year can be as helpful as I would estimate $5,000 of advertising). Your NOC people should work closely with cybersecurity, server ops, and DevOps.
But here’s a heretical take: I’m not convinced that flooding the NOC with junior engineers is the answer. Training is good, but in touchy environments —like the ones I worked at in the banks — experience beats eagerness all day long every day.
Automation is necessary — provided it’s done right. You don’t want to be automated simply for the sake of automation. I don’t like it when vendors decorate their product brochures with AI-powered without content. Here’s the thing about AI in security: it’s more like a hype meal that’s promised time and time again to taste like something gourmet but then delivers the same old microwaved leftovers.
Pay attention to automation that does:
Do not automate the decisions that require human judgment — at least, not yet. Believe me, when I watched an automated reply accidentally turn off mission-critical bank systems at 3 a.m., I cringed.
Scattering your tools is the worst. You tack on another monitoring platform here, a fresh SIEM there, you throw in an endpoint detection system and soon, your NOC analysts are juggling tools far more than solving problems.
What I’ve found works best:
Oh, and here’s a bit of a flashback – once upon a time we even had single box solutions that also handled voice and data multiplexing over PSTN lines. The hardware was clunky; the interfaces, rudimentary — but it worked. Recently, our tools must be faster, more intelligent, but also simpler to use. Or else we’re simply reproducing all these same problems dressed up nice.
At PJ Networks, we don’t do one-size-fits-all. It’s these banks we joined forces with – three fast-expanding banks whose NOC were on their knees because of alert storms and partitioned teams. Here’s what we did:
The result? An NOC that matched the growth of their business, never skipping a beat.
The network landscape is constantly changing — 5G, edge computing, IoT — and your NOC needs to be prepared. Zero-trust isn’t just a buzzword anymore (and I’ve actually worked with 3 banks on that — some can do it better than others). Your NOC needs to shift from a reactive unit into a proactive force for securing the business.
That means:
One thing is clear: No technology will make up for not having skilled staff and strong processes.
Scaling your NOC for a fast-growing business is no piece of cake—more like trying to tune a supercar’s engine while the engine is racing. Together with the right mix of experienced individuals, intelligent automation, and enlightened tool selection, you can create a contender for the network operations title.
We’re fortunate to work with clients facing these challenges every day at PJ Networks. If you’re a CTO or IT director grappling with every one of these growing pains, take my word for it: plan well, hire smart, and don’t be seduced by every shiny new tool rumored to be a panacea.
And hey — let’s never underestimate the importance of a great cup of coffee when you’re up late juggling network alerts!
P.S. Not all password policies are created equal, and that is a good thing. If there’s one rant I will always have, it’s password policies that feel the need to excommunicate users through complexity rules that don’t even make you more secure. But that’s perhaps a story for another blog post!