Newsroom · Perspective
Design the limits first
When people picture automation, they picture what it does. The more important question is what it refuses to do.
On FYND, the platform we built to connect modeling agencies with talent who chose to be found, the most important features are the things it will not do.
The risk was in the outreach
Agencies scout people across social platforms, and every lead has to be tracked: who was contacted, who replied, who is on the roster. Automation can take most of that work off a team. But outreach carries real risk. People under 18 must never be contacted. People must be able to opt out. Email volume needs limits. A tool that ignores these puts the agency at risk.
So we designed the limits first
- Under-18 screening. Profiles that look under 18 are hidden from agency search, and a card cannot go live without a confirmed 18 or over.
- Privacy by default. The talent's email address is never shown to the agency.
- Outreach with limits. Emails go through an outbox, follow-ups are scheduled automatically, and unsubscribes and replies are tracked.
The routine work runs by itself. The limits are what make it safe to let it.
Have a process that carries real risk?
A discovery call is a conversation about what you are working toward. We start with the limits, then the automation.
The principle
Good automation is mostly restraint. Decide first what the system must never do, then build everything else inside those limits.
That applies to any process where a mistake would hurt someone. Before you ask what to automate, ask what must never happen.
Our AI pilot on one page has a line for exactly this: what stays human, and what information may never be used.
FAQ
Frequently asked questions
What does "design the limits first" mean?
It means deciding what a system must never do before deciding what it should do, and building those limits into the system rather than relying on people to remember them.
Which processes need this most?
Any where a mistake could harm a person or break a rule, such as contacting people, handling personal information or moving money.
Does adding limits make automation less useful?
No. Limits are what allow you to trust it with routine work, so the routine work can run without constant supervision.
Where can I see the example?
On the FYND case study, which describes the platform and the safeguards built into it.
