Last time I told you I've been through a platform shift before, and that I think AI is the next one.
This time; where I think the opportunity actually sits. Hint: I don't think it's where everyone is looking.
Worth saying what this isn't about. ServiceNow are doing a good job of building AI into the platform; Now Assist, AI Agents, the Autonomous Workforce. That story is well covered, and it's heading in the right direction.
But there's one thing that jumps out at me, waving. Platform AI executes work inside workflows. It doesn't sit in your discovery workshops. It doesn't turn three hours of app-owner conversation into a service model. It doesn't check a data modelling workbook and spot the handful of relationships that would silently fail on import. It doesn't translate your project plan from geek into English when the delivery team tells you (politely) that it's daunting.
That's delivery work. It's the work that decides whether an implementation succeeds. And when it comes to applying AI to it, almost nobody is showing their working.
There's a second part to this. Anyone who's been around the platform for years is sitting on a pile of delivery knowledge; solution designs, lessons learned, the patterns you carry in your head and never quite get time to write down. Practically none of it is being consumed by AI. It sits undisturbed in SharePoint, local disks and people's memories while every new engagement starts from a blank page.
Just think about that for a few seconds. How many years of knowledge and experience are buried in your company SharePoint right now? How do you tap into that collective memory?
My thinking is, use that knowledge to build AI capabilities, point them at the delivery work around the platform, and you have an instant force multiplier for your entire delivery team. I've been spending my evenings and weekends experimenting with this for months now. I'm not exaggerating when I say the potential here is beyond epic (excuse the hyperbole, this stuff is genuinely exciting).
It's not flawless. You can't trust it completely, and you certainly can't leave it unattended to churn stuff out for you. This is not consultancy on autopilot by any stretch of the imagination. But I'm still blown away, and I'll show you the receipts as we go. Nothing I publish goes out on a single pass; everything gets verified at least twice, once by an AI reviewer wearing a deliberately hostile hat, then by a completely different vendor's model, then by me, because my name is on it. Belt, braces and superglue.
I don't have a tidy name for this space, so I've been calling it ServiceNow AI Adjacent. AI augmented delivery, AI assisted delivery, AI consultant assistance; call it what you like. Everything around the platform rather than in it. Implementation, workshops, data modelling, planning, governance, handover.
Next up, how it started. A hectic project, a sweet shop, and the plonker of the year award.
If you're working in this space too, or you think I've wandered off down a blind alley here, say so on the LinkedIn post. I'd rather find out now.
First published on LinkedIn, 7 July 2026. The conversation lives there; the writing lives here.