Your AI panel of experts (and how to build one in 10 minutes)
What if every piece of work you produced got reviewed by a panel of experts before anyone else saw it?
A tax specialist. A grant assessor. A contract risk reviewer. A plain-language editor. All on demand. No billing rate. No waiting.
Last week I worked with an Australian research institute. We worked on AI strategy, tools, automations, and risk. But we also spent a lot of time with hundreds of people on prompting techniques that genuinely improve workflows.
This is one of them. And while the example comes from research, the principle applies to every profession. Including yours.
The idea is simple: give AI a specific identity.
Tell it who to be. What they know. How they respond. Then give it your work to review.
That's called a persona prompt. And what you get back is not a generic AI response. It's feedback from someone who actually knows what they're looking for.
Let me show you what I mean.
The grant reviewer from NHMRC (National Health and Medical Research Council)
I built a grant reviewer persona and pasted it into Copilot, along with a grant abstract. Here's the prompt I used:
"You are a senior grant reviewer for the NHMRC with 20 years of experience. You are direct, rigorous, and you do not soften feedback. Review the following grant abstract for clarity, evidence gaps, and likelihood of funding success. Highlight anything a reviewer would flag."
What came back was not gentle. The AI assessed it as a senior panel member on a competitive national scheme. It said the abstract was generic to the point of interchangeability. The rationale was absent. Timing inconsistent with the trial length. Likelihood of funding success: frankly very low.
Every point was correct. The researchers in the room recognised every flag.
The reviewer was right on every point.
So we rewrote the abstract. Came back to the reviewer. The feedback was much better. And still there were suggestions.
So it’s not a magic prompt. It’s iterations.
Three rounds of rigorous review before the application goes anywhere near a real reviewer.
What does this mean for you?
You may not be writing research abstracts.
But think about where in your work you could use a reviewer like this.
Reviewing tax returns. Reviewing contracts. Reviewing a client letter before it goes out. I have agents and skills that review my workshops, my writing, my executive briefings, my ideas.
Here are two to get you started:
"You are a senior tax accountant with 20 years of experience reviewing individual and small business income tax returns in Australia. You are thorough and direct. Review this tax return for errors, missing deductions, inconsistencies, or anything the ATO might query. Flag each issue clearly."
"You are a senior commercial lawyer with 20 years of experience reviewing contracts for small to medium businesses. You are thorough and direct. Review this agreement and flag anything unusual, risky, or missing standard protections."
A medical practice manager reviewing patient communication. An HR adviser checking a termination letter. A financial planner running an advice document past a compliance reviewer. Any expert. Any domain. One prompt.
Your best experts might not be external at all. Your most difficult client. A junior who's never seen this before. You can rehearse any audience before you walk into the room.
Iterate.
You don’t need to get the perfect prompt on day one.
I never do.
I iterate. I update the prompt over time. Add examples, add context, add the things you always need to check or to correct.
Once I had a good conversation with AI and landed somewhere useful, I ask it to write me a reusable prompt for next time. Sometimes I record myself talking through how I complete a task, transcribe it, and turn that into the prompt.
Reuse.
In Copilot, save it as an agent (from Copilot Chat).
Give it a name, a description, paste in your persona. Next time, open the agent, paste in your work, hit send. Done. Or even easier, mention the agent half way through a chat.
In Claude, the same principle applies with Skills. Build the persona once, save it as a Skill, and it's a mention away in every session.
In ChatGPT, create a custom GPT. In Gemini, use Gems. Same idea across all of them. Build it once. Reuse it every time.
At this point, we still want everything to go through human review too. And that’s where the majority of the time saving happens. Because you’ve done enough iterations to get it really good before taking up the time of a senior reviewer.
One prompt. One expert. On demand.
Build a panel. Run your work past all of them before anyone else sees it.
That's your AI panel of experts.
Want to work on this together? Come find me at inbal.com.au/events.
—
Inbal Rodnay
Guiding Firms in AI Adoption and Automation
Keynote speaker | AI Workshops | Executive briefings | The Tech Savvy Firm
Want to receive these updates straight to your inbox? Click here: www.inbal.com.au/join
When you are ready, here is how Inbal can help:
AI Agents in Accounting | Hands-on Workshop (Online, 24 July)
- Understand what AI agents can realistically do in accounting workflows, and how to design safe, sensible use cases.
- Build several working agents step-by-step and leave with reusable patterns for your practice or firm.
4 hours, online, interactive with breaks and Q&A

