You can build a dashboard. It's easier than you think.
You can build an interactive dashboard with AI in just a few minutes. It's not hard.
Once you get started, you’ll find so many uses for it.
All you have to do is ask.
People are building dashboards for lodgement status, for job tracking, or even entire financial and operational dashboards for their clients.
And I often build one when I have a big task with 25 different things to do, I find it overwhelming, and I just want something visual where I can tick them off as I go and get confetti every so often.
So let me show you how I do them.
In the video I show it in Claude, in Copilot chat and in Copilot Cowork.
ChatGPT is just the same.
Make me an HTML dashboard
Here's is my prompt:
"This is my firm's credit card statement. Make us an HTML dashboard of where our money went, with our software subscriptions listed separately."
That's it. Make me an HTML dashboard. These are the magic words.
I gave it a folder to work in, and it went off and built it.
It even found my brand kit skill (remember we talked about skills the other week?), so it used my favourite colours.
It showed spend by category, the software subscriptions listed separately, and all the transactions underneath.
That's how easy it is.
Now I can iterate and ask: Add an option to drill down, click for more information, add filters, sorting or some analysis on the page, I just ask.
And in Copilot
I did it twice in Copilot: in Copilot Chat and in Copilot Cowork.
In Copilot Chat I put in exactly the same prompt. It created the dashboard, I opened it, it look a bit broken. So I took a screenshot and said, "This is what it looks like." And it corrected it and gave me a better one.
Copilot Cowork was my winner for the day.
Same prompt, I pointed it at a OneDrive folder, and it did the job.
What I really like is that it also added intelligent suggestions on how to improve our spend.
The difference between the two?
Chat gave me something to download
Cowork saved it straight in the folder for me, and it thought deeper and came back with ideas.
Then I scheduled it
After I looked at the dashboard, I asked Cowork: "Schedule this task so it checks the folder every day at 12pm, and if there is a statement, it updates the dashboard."
And it scheduled it for me.
When I ran it as a test, it also sent me a Teams message: "Hey, there's a new dashboard for you. Here's the link to open it."
We know Copilot Cowork has a cost a bit every time we use it, and I like checking with the cost command. The first dashboard was 55 US cents. Setting up the schedule was another 87 cents. I imagine each run from here will be some 50-70 cents every time. (This was a small statement.)
A thing to remember
I've seen people in the club create dashboards for internal use and also for their clients.
If you're thinking about giving your clients access to the dashboard, we need to consider where we store the data, how we provide secure access, and more.
But if you're using it internally, the risk is pretty low.
You're just organising data to look in nice ways.
And you can really impress your client with that.
Start with the data you have
The thing that can sometimes slow us down is getting the data.
In this example I downloaded a statement.
If I wanted to automate it further, I could automate the downloading. I could use APIs or MCPs to get data from different sources.
That’s the advanced class.
To begin with, just build yourself one simple dashboard, so you have the hands-on experience that you can do it.
Go ahead and ask your favourite AI: "Make me an HTML dashboard".
Then iterate to add the interactivity you want.
What are you going to build?
--
Inbal Rodnay
Guiding Firms in AI Adoption and Automation
Keynote speaker | AI Workshops | Executive briefings | The Tech Savvy Firm
When you are ready, here is how Inbal can help:

