Vigil is continuous operational data assurance. It takes the metric definitions you already proved out, OEE, throughput, scrap rate, on-time, unit cost, and re-checks them every cycle against live data. Not the whole raw feed. The specific numbers a leader would act on.
When a definition holds, Vigil says so and stays quiet. When one moves, it tells you which metric, by how much, and what changed underneath: which table, which code, which default. The same inspectable detail Validate produces, except now it never stops running.
This is not monitoring and it is not support. Uptime tools watch whether the system is running. Vigil watches whether the numbers still mean what you decided they meant. That is a different question, and it is the one that costs you when the answer is no.
Every check has an expiration date nobody prints on it. The day a Verify or a Validate engagement ends, the assurance ends with it. The numbers were sound at that moment. Nothing holds them there.
Drift is quiet by nature. No error, no failed load, no alarm. A number keeps rendering, it keeps looking plausible, and it’s wrong for weeks before anyone catches it, usually in a meeting, usually in front of the person who acted on it. By then it isn’t a data problem. It’s a dispute about whose number to believe.
Trust decays silently. Vigil is how it doesn’t.
A watch that catches the shift the week it happens turns a boardroom argument back into a five-minute fix. That is the whole trade: a small standing cost against the price of finding out late.
- Lock the definitions. Start from the metrics already pinned in a Verify or Validate engagement, so the watch begins from numbers that were proven, not assumed.
- Re-check on a schedule you set. Vigil re-runs those definitions against live data on the cadence you choose. It runs on hardware you own, against data that never leaves your control.
- Stay quiet until something moves. No noise when the numbers hold. When one drifts past the tolerance you set, Vigil raises it, with the metric, the size of the move, and the change that caused it.
- Show the work, every time. Each result is inspectable down to the definition and the values behind it. You act on the finding because you can see exactly what produced it.
A company that can trust its operational numbers continuously stops relitigating them and starts deciding from them. That is the rung Vigil sets you on.
The OIM is what you build once you’re there: the operations intelligence layer your company owns outright, run on numbers you’ve already watched hold. The same people who verified the migration, validated the move, and kept the watch, now building the model you keep.