Long title, eh?
If you work in tech long enough, you eventually face that humbling moment where you realise you've become a beginner again. I wrote about this recently on LinkedIn, the whole unlearning and reinventing-yourself theme. When I transitioned from ERP to data solutions about four years ago, I had the luxury of doing it safely, with the support of my manager back then, Anish Shah.
My "dunce" moment arrived the day I stepped into the world of ServiceNow, a powerful platform in orchestrating enterprise workflows, yet can be complex enough that your first login screen feels like being dropped straight into a cockpit. I often joke with colleagues who also come from command-line or coding-heavy backgrounds: on text-only platforms, you can work with cryptic switches, understand how the commands work, and reproduce results fairly quickly. On UI-heavy platforms, the locations of buttons feel almost more arbitrary. But let’s not get stuck on the UI, the heart of ServiceNow is still the engine beneath it.
After years in SAP, analytics, data science, architecture, and presales, I am sharing the story of my CSA journey.
The “I’ll pretend I know what this is, but I’m too afraid to ask” phase
There I was, nodding confidently while running an internal monologue:
“Tables and forms and ACLs.
What is even a context menu?Everyone else seems to get this.Keep nodding. Don’t blink. Maybe blink twice… internally.”
If you’ve ever switched ecosystems, you know this phase well.
The fear is irrational, but very, very real.
And that meme became my spirit animal during the first few weeks. My manager, Rajat Rakesh, encouraged me that while the "wall" described in the meme is real, getting over or through it is just as real.
Where I really started
Instead of reading everything in order (which my brain refuses to do), I jumped straight into building. That meant:
-
constructing early POCs and breaking them
-
fixing them and breaking them again
-
realising the platform rewards curiosity far more than memorisation
The more I built, the more things clicked: data models, user roles, flow logic, the "ServiceNow way" of structuring work. Writing about it publicly helped solidify concepts and forced me to stop hand-waving explanations… although I’ve been told a few times that I over-gesticulate. That is not going to change. Not now. Not ever.
How I prepared for the CSA (no secrets, no exam content, no spoilers)
I'll avoid anything even remotely resembling exam content. Here’s what actually helped:
1. Doing real work on the platform
Nothing teaches you faster than debugging your own configurations (and questioning your life choices in the process).
2. Understanding concepts, not questions
Instead of memorising anything, I focused on why things behave the way they do.
3. Reading documentation… selectively
You don’t need to memorise docs.
You just need to know where the answers usually live.
You just need to know where the answers usually live.
4. Repetition disguised as POCs
Every demo, every flow, every AI Agent I built became accidental study material.
5. Not ignoring the Inkling eBook
It’s tempting to breeze through the course videos and call it a day. Don’t. The Inkling eBook is key, treat each exercise with respect.
Remember the names of buttons, panels, and elements. Yes, it feels forced, but the world is a better place when we all agree on what a context menu is.
Remember the names of buttons, panels, and elements. Yes, it feels forced, but the world is a better place when we all agree on what a context menu is.
Why the CSA still matters in the age of AI Agents
There’s a misconception that AI Agents will eventually make platform knowledge optional. My view?
AI makes fundamentals more important, not less.
Why?
1. AI Agents rely on your platform’s foundation
If your tables, relationships, and access rules are messy, AI Agents will automate the mess faster.
2. AI amplifies everything: the good and the mistakes
A human misconfiguration is an inconvenience.
An AI-automated misconfiguration is a business incident.
An AI-automated misconfiguration is a business incident.
3. The future isn't "AI or builder." It's "AI-accelerated builder."
AI helps you build faster, not understand deeper.
That deeper understanding is still very much CSA-level, and beyond.
That deeper understanding is still very much CSA-level, and beyond.
4. You can't govern what you don't understand
Even the most advanced agent needs guardrails.
Those guardrails come from human intent, platform understanding, and architectural judgment.
Those guardrails come from human intent, platform understanding, and architectural judgment.
CSA gives you exactly that.
The final output variable (or is it?): the CSA wasn’t a badge, it was the on-ramp
Looking back, the CSA didn’t mark the end of learning. It marked the beginning of clarity:
-
clarity in how the platform thinks
-
clarity in how data flows
-
clarity in how AI Agents should be designed responsibly
It also marked the moment I graduated from the “I’m too afraid to ask” phase into the “Okay, now I can ask better questions” phase.
If you’re just starting or feeling intimidated, remember: every expert you admire had this exact meme moment. Learning something new requires equal parts curiosity, humility, and humour.
Lastly, a conversation with Sean McFarlane inspired me to write this as he steps into his new role. As a team, when something is new, it’s more fun if we learn the gnarliest problems together. For me, the easiest way to start anything unfamiliar is to tell myself:
“I have no idea what's going on… but I’m going to learn it anyway.”
