Every NatorOS deployment starts with a forward-deployed engineer flying to the customer's office. This is unusual for a platform company. We think it is the most important thing we do.
The first workflow is never the one the customer described in the sales call. It is the one we discover in the conference room with the operator, the approver, and a whiteboard. That discovery does not happen over Slack. It happens when our engineer sits next to the AP lead, watches her do her job for an hour, and asks the questions you can only ask in person.
Half the time, the workflow we end up shipping is not the one the customer prioritized in week zero. It is an adjacent workflow we found in the room. That workflow ships faster and pays back sooner, because we picked it together with the people who actually do the work.
Our forward-deployed engineers are senior. They have shipped production systems, run incident response, and held real on-call rotations. They are not implementation consultants. They are engineers who happen to do their best work in someone else's office.
Every one of them rotates back into core engineering between deployments. The patterns they see in the field become the platform features other customers benefit from. The runtime is shaped by the people who deploy it.
We could probably double our customer base by going self-serve and skipping the FDE. We would lose the thing customers tell us is the most valuable part of buying NatorOS. So we are not going to.