What I actually do all day, as an AI consultant (honest edition)
The first thing I open every morning is email. Then Slack. Then ClickUp. Here’s what that actually looks like.
The morning
Email, then Slack, then ClickUp. In that order, every day.
I have a ClickUp superagent set up that pulls my project timelines together and gives me a clear picture of where everything sits across all my clients. That alone saves me 20 minutes of digging around trying to remember where I left things.
Then I open a Zapier SDK app I built myself. It generates a timeline with a checklist of my movements through the day: emails, meetings, project tasks, personal commitments, in the order they need to happen. Nothing off the shelf did exactly that, so I built it.
By 9am I know what the day is. Not what I planned. What it actually is.
What I actually spend most of my time on
About 60% of my time is building. Automations in Zapier or n8n, AI workflows in Claude, tools for clients, internal systems. This is the core of the work and it’s where most of the day goes.
20% is meetings and follow-ups. Discovery calls, check-ins, proposals, emails. The communication layer that keeps everything moving.
The remaining 20% is teaching, training, and workshops: helping client teams actually understand what they’ve got and how to use it.
That last bit surprises people. They assume AI consulting is mostly the technical build. It is, mostly. But the teaching sits alongside it at every stage, and it’s what determines whether the build actually sticks.
What deep work looks like day to day
Most of my focused work time goes into four things.
Building automations in Zapier or n8n. Building out client-facing tools and apps in Lovable. Thinking through workflow design in ClickUp. Recording Loom walkthroughs so clients can refer back to things without needing me on a call. And editing Claude outputs before they go anywhere near a client.
That last one surprises people. The model produces a draft, an email, a summary, a document. I shape it. I bring in the context the model doesn’t have: the client’s tone, their specific situation, what their team will actually read and respond to. This is where a lot of the craft sits, and it’s one of my favourite parts of the day. Taking something rough and making it genuinely useful.
The tools I come back to every day
Claude for thinking and writing. Zapier for automation. ClickUp for client work and project management. Loom for everything that’s faster to show than explain.
I test a lot of other things. Those four are where most of the real work happens.
How the week flows
Early in the week I plan, research, and catch up on what’s changed in AI since the previous week. Which is always something. The pace of change is absolutely mental right now and staying current is part of the job.
Midweek is for clients: calls, workshops, implementation sessions. These are the days that look most like what people picture when they imagine AI consulting.
By Thursday and Friday I’m in documentation and handover mode. Making sure everything is clearly explained, properly recorded, and something a client can actually use without me being on a call every time they need it.
What the day feels like
Honestly? Chaotic but satisfying in a way I genuinely didn’t expect when I started.
Something rarely goes exactly to plan. A tool behaves unexpectedly. A client has a new requirement that reshapes the approach. You adapt, figure it out, move forward. And at the end of the day there’s usually something that didn’t exist in the morning: an automation running, a process documented, a team that can now do something they couldn’t do before.
That’s the bit that gets me. The tangible “this is working now” feeling.
The part that matters most
AI consulting in 2026 is really about closing the gap between what AI can do and what a real business will actually adopt and use. That takes enough technical skill to build things properly, but the bigger requirement is the human side: listening carefully, translating between what’s possible and what’s practical, and caring genuinely about whether the thing you’ve built is actually making someone’s working day better.
The tools are getting easier every month. The human side of implementation is not.
→ That’s where the work lives, and it’s also why this role is one of the most interesting ones to be in right now.
In short: the job is more hands-on, more people-focused, and more genuinely varied than the outside view of it suggests. It involves Slack at 7am and editing AI outputs and spending more time on adoption than on anything technically flashy. I wouldn’t swap it. There’s something brilliant about working right at the edge of what’s possible and helping people actually get there.