A public report on the Upwork automation market for Maker School's members. It collects its own data every day, then writes, checks and publishes itself on the first of each month.
Continued from work originally published by Aaron Melton at Ascend Automation Agency.
The report started as Aaron Melton's. For twelve months he exported the data, ran his script, wrote two reports, updated the dashboard by hand and posted it for the community.
When he stepped back, we picked it up for Maker School and rebuilt it to run on its own.
New Upwork job posts are collected every day, with no export and no script to run.
On the first of the month it runs the numbers, drafts both reports and checks every claim against the data.
A sentence that contradicts the numbers blocks the publish until it's rewritten.
Monthly pages, a 12-month view, RSS, and the data as CSV or JSON.
This is what we built for Maker School. The demo above is similar software on invented data, not their system.
The software does part of the work at 5 of 5 steps. A person still reviews the community post before it goes up.
The report now covers more than 70,000 Upwork job posts: 53,786 across Aaron's twelve months, plus more than 19,500 collected since May. Five monthly editions so far, from May to September 2026, with the numbers tested to within 2% of the original method.
Proven services underneath, so the report keeps running on the first of every month without anyone starting it.
About the clientNick Saraev's community for people learning to build and sell AI automation.
Visit skool.com/makerschool ↗We'll make it collect its own data, write itself and check its own claims, with a person signing off.
Start with an AI audit. We'll map which system to build first.
