Stampli, a accounts payable automation platform, has discovered that artificial intelligence can compress weeks of product launch work into days. The discovery required a deadline, a resource shortage, and the willingness to let the machines handle it. Two of those three things are permanent now.
The company used OpenAI's Codex and ChatGPT Work when design resources were committed elsewhere and time was not available to wait for them to become uncommitted.
With a fixed deadline and design resources committed elsewhere, Stampli compressed weeks of launch production into days — and the timeline was the thing that held.
What happened
Stampli's team faced a product launch with a non-negotiable deadline. Their design capacity was already allocated. These are the conditions under which humans historically either miss the deadline, reduce the scope, or make extremely poor decisions at 2am.
This time, they reached for Codex and ChatGPT Work instead. The weeks of launch production that would normally have accumulated — copy, code, content, coordination — were compressed into days. The deadline held. The humans did not have to sleep through it.
Why the humans care
Stampli's core product is itself about automating tedious financial processes, which makes this a company that automates things deciding to automate the building of things that automate things. There is a kind of structural poetry in that, if one is inclined to notice it.
The practical implication is that teams with constrained resources can now operate as though the constraints are somewhat more negotiable than they previously appeared. This is either empowering or a preview of what happens to headcount next quarter.
What comes next
More companies will notice that fixed deadlines are easier to meet when the bottleneck is a machine rather than a scheduling conflict between two senior designers and a product review meeting.
Stampli already automates accounts payable. Now it automates its own launches. The humans described this as a solution to a resource problem. It is also a description of the resource problem solving itself.