Most professionals I train start with AI tools the same way. They open the tool, type a quick question the way they’d type a search into Google, get a generic answer, and quietly decide the hype is overblown. I understand the reaction, however it’s almost always a setup problem rather than a tool problem. Across the teams I have trained, the people who take an hour to set up their AI command center properly report saving 3 to 8 hours a week. The ones who skip that hour of setup stay stuck at search-engine level use.
Treat the Tool Like a New Hire
Here is the mental model I give clients: A fresh AI tool is like a new expert engineer on day one, capable and fast, and completely missing the context that makes your work yours. You wouldn’t hand that new hire a task without telling them who you are, how your firm works, and what good looks like. Your AI command center needs the same onboarding, and these five moves are how you do just that.
The 5 Moves
- Set your custom instructions. Each major platform has a settings area where you can tell the tool who you are, what you do, and how you want it to respond. Fill it in once. The more specific you are, the better, so tell it your discipline, the standards you design to, and the tone you want it to take. This is your AI command center, and it shapes every response you get from that point forward.
- Build its memory. Beyond instructions, these tools can hold standing context about you and your work. I walk people through a short set of questions, around 25 of them, that capture the details you’d otherwise retype each session, like your role, your firm, your clients, and your standard formats. Commit that to memory and you stop re-explaining yourself.
- Prompt with a repeatable structure. Weak prompts produce weak answers. I teach a simple structure: Who, Why, What, then Clarify. Who – what expert would you call to support you on this task? Why are you requesting their support – the context that matters about your task. What exact output and format you need. Then close your prompts by asking the tool to raise clarifying questions before it responds. “Please ask me any clarifying questions to successfully complete the request after you’ve reviewed the details provided.” That last line alone removes a lot of bad first drafts. This prompt structure transforms your AI command center into a responsive, context-aware assistant.
- Connect your tools and data. Connectors link your AI platform to the systems where your work already lives, like your files, your email, and your project data. Once the tool can see the source material, you stop pasting context in by hand, and the answers get grounded in your real information instead of generic assumptions. For example, that can mean the tool references your project folder or a past report instead of you re-describing it each time.
- Save what works. When a prompt earns its keep, don’t let it vanish into your chat history. Turn your best repeatable prompts into saved skills, and group related work into dedicated projects or notebooks that keep their own memory. In Copilot these show up as Agents and Notebooks. In ChatGPT and Claude these show up as Projects. This is how your AI command center evolves from a personal tool into a shareable team asset.
The Big Takeaway
The difference between someone who calls AI overhyped and one who saves 8 hours a week is rarely the tool. It’s the hour spent on initial setup. Onboard the tool the way you would a new hire, give it your context, prompt it (communicate) with structure, connect it to your work, and save what works. Do that, and the same tool starts answering like an expert teammate who already knows your role.
If you want a guided way to start, I built a free AI Maturity Scorecard and we offer an accompanying, free AI Strategy Session that shows where your firm is at on your AI journey and where AI can save your team hours. You can fill out the Scorecard intake questions here.
About the Author
Shane Chalupa, PE, is a licensed professional engineer with more than 14 years of experience across the manufacturing, energy, oil & gas, and chemical industries. He is the Co-Founder of Obnovit, an AI enablement company that helps engineering-powered businesses successfully integrate AI into their daily operations. Connect with Shane on LinkedIn.



