There are three steps, and none of them cost you anything.
You describe the request.
A paragraph is plenty. Tell us what happens today, whose time it costs, and what a solved version would look like.
We select a few and build.
We choose the requests that will teach the most people something, and we build a working solution against realistic sample data that we construct ourselves.
We publish the whole thing.
You get a short video, the prompts, and the file, and they are free to take and reuse. You decide whether you are named, kept anonymous, or left out of it entirely.
Everyone who writes gets a reply, whether or not their request is selected to be published. This information is provided for educational purposes: any solution provided is a draft for your team to review and adopt per your company’s internal AI policy, and OAS does not take on management responsibility for it.
The best submissions are the most challenging and specific.
Not “how do we use AI.” Name the work itself, and the harder and more particular it is, the better. For example:
“Our monthly variance narrative is the problem. The numbers are ready by day three, but the write-up still takes a full day.”
“We receive forty-page vendor agreements, and we have to find the five things in each one that actually matter to us.”
“Our board packet contains the same twelve exhibits every month, and we rebuild every one of them by hand.”
“We need a policy that nobody has time to start, and we face a blank page every time we try.”