I don’t have one AI assistant. I have AI employees.
By Zahra Ibrahim · 3 min read
The shift that made AI genuinely useful in my work wasn’t finding one perfect prompt. It was giving different chats different jobs.
For a long time, I treated AI like one endlessly patient assistant. I opened a chat, asked a question, copied the answer and moved on. It was useful but it still felt like a tool I had to manage from scratch every time. I would end up with multiple segmented chats, one little thing here and there. A good start but certainly not optimised for a growing workload!
Then I started separating my chats. One AI space for each client. Another for SOPs. Another for content. Another for the business I’m building. Each one held the right background, the right language and the right kind of decisions. In other words, projects.
I do want to stress, though, that my projects literally just started as chats. Sometimes you won’t have enough material to feed into a project and that’s ok. If you use ChatGPT you can also start with a normal chat and transfer this into a project .
Level one: a new chat every time
This is where most people begin. You explain the task, add the context and hope the output is good enough. It works for isolated questions, but the setup repeats and the quality depends on what you remembered to include that day.
Level two: a dedicated AI for each client
This changed my day-to-day operations work. A client-specific space can learn the formats repeatedly used, the systems in their business, the names of their offers and the difference between how they speak internally and externally.
It becomes less like briefing a stranger and more like returning to a colleague who understands the account. I still review everything, so the judgment stays with me but I no longer rebuild the context before every task.
Level three: AI inside the system
The next step is connecting that thinking to the work itself: meeting notes that become tasks, voice notes that become end-of-week updates, or an onboarding workflow that prepares the right communication at the right stage within your Dubsado.
This is where AI stops being another tab and starts becoming part of your business operations.
The point isn’t to replace your brain
I don’t want AI making every decision for me, and neither would you. I want it holding context, reducing repetition and helping me get to the useful part of the work faster.
The best AI setup is not the one with the most impressive automation. It’s what’s actually workable for your business needs at any time. Automations can be changed around, added to or reduced, so nothing really is fixed.