What you will find here
Every playbook in this category describes one workflow end to end: the business problem, what
the manual process looks like today, the automated version step by step, what it takes to
implement, and where it needs oversight.
The steps are labelled by who performs them — Human, AI or System — because that labelling is
the whole argument. Automation that quietly moves a judgement call from a person to a model,
without anyone deciding to, is how these projects lose the trust of the team that has to run
them.
How to read a playbook
Start with the workflow. If the steps marked Human are the ones your team would refuse to
give up, the scope is right. If they read as rubber-stamping, what you are looking at is an
approval queue with extra steps, not automation worth paying for.
Nothing here is a product. Each playbook is a design we have thought through and would build
against; the numbers attached to it are ranges that move with your volume, not results we are
claiming on your behalf.