OPERATIONS CREW
Your company shouldn’t need you to check that everything is working.
A process exists. A task is assigned. Everyone thinks it’s handled. Until you discover it wasn’t. Your Operations Crew watches how the company actually runs, spots things that are slipping and helps resolve problems before they become repeated failures.
01 · THE PROBLEM
A process isn’t working just because it exists.
You can have documented workflows, controls and systems, yet still depend on somebody noticing when they stop working. That somebody is usually the founder.
01 · MONDAY
Everyone thought it was handled.
What slips
A new customer signs up. The handoff to onboarding is documented. Nobody confirms it actually happened.
02 · WEEK THREE
The check that quietly stopped.
What slips
A weekly reconciliation stopped running after a tool changed. Nothing failed loudly, so nobody noticed.
03 · MONTH TWO
The same failure, again.
What slips
It was fixed once. Nobody checked the fix held, so you hear about it from a customer.
02 · GIVE THIS RESPONSIBILITY TO CREW ONE
Give this responsibility to Crew One.
Carries the watching, so you step in only when a decision is needed.
Job
What Operations carries
Off your plate
- 01
Watch critical processes
The recurring responsibilities your company relies on, and the result each one is meant to produce.
Off your plate: Chasing whether things actually happened.
- 02
Spot exceptions
Failures, delays, missing handoffs and controls that have quietly stopped working.
Off your plate: Being the one who eventually notices.
- 03
Recommend repairs
What changed, the likely cause and what to do next, set out for your decision.
Off your plate: Diagnosing every failure yourself.
- 04
Follow up on fixes
Corrective actions followed up, with a watch kept for the same failure returning.
Off your plate: Hoping the fix held.
03 · THE OPERATING LOOP
From drift to repair.
Compares how the company should run with how it actually does, and brings you in when a repair needs approving.
- 01
Declared operating process
What should happen, who owns it and the result to expect.
- 02
Observe actual behaviour
How the process really runs, run by run.
- 03
Compare expected vs actual
The expected result set against what happened.
- 04
Detect failure or drift
Failures, delays, missed handoffs and slow drift.
- 05
Recommend corrective action
Evidence, likely cause and the proposed repair.
- You decide06
Governed repair
You approve the repair before anything changes.
- 07
Watch the effect
Did the repair produce the expected result?
- 08
Watch for repeat failure
The same pattern is watched for on every run.
Back to observing
The loop continues on the next run.
Where your judgement comes in
CARRIED WITHIN AGREED AUTHORITY
Every outcome returns to the start
04 · WHAT THIS LOOKS LIKE
What this looks like.
Not another alert. A record of what was expected, what actually happened, what to do about it and what happened next.
◇ OPERATIONS · EXCEPTION
PreviewAn operating process has stopped working.
OPENED 08:10
- 01
Customer onboarding
Two handoffs have not been confirmed.
Recommended
Investigate the missing transition before processing more customers.
Inspect evidence
OPERATING RECORD · CUSTOMER ONBOARDING HANDOFF
Customer onboarding
Evidence
EVIDENCE · HANDOFF TRAIL
Two completed signups have no confirmed owner. Both followed a change to the signup form.
- Customer A
- Signup form
- Customer B
- Customer C
Judgement
The change most likely broke the transition. Restore it, then reassign both customers.
Your approval
RECOMMENDATION
Investigate the missing transition before processing more customers.
Waiting for you. No process changes until you approve them.
Outcome
Watch the next handoff
After correction, the next handoff and the pattern are watched.
Watching continues within agreed authority. Any change to the process waits for your approval.
Illustrative workflow. Early-access capabilities and integrations are confirmed during onboarding.
05 · YOU KEEP CONTROL
You keep control.
Operations carries the watching and the diagnosis. Every repair, and every change to how the company runs, comes back to you.
CARRIED WITHIN AGREED AUTHORITY
- Watching declared processes against their expected results
- Comparing expected with actual behaviour
- Flagging failures, delays and missing handoffs
- Setting out evidence, likely cause and a recommended repair
- Following a repair through to its result
Comes back to you
- Approving any corrective action
- Changes to a process, an owner or a control
- Anything that affects customers, money or people
- Deciding when a repeated failure needs a different design
The boundary is agreed with you during onboarding. Anything outside it comes back to you first.
07 · QUESTIONS, ANSWERED
Before you hand it over.
Clear processes. Agreed authority. Your final say.
- 01
Will it change our processes on its own?
No. It recommends corrective action and brings it to you. Repairs go through a governed step that you approve.
- 02
What does it need to watch a process?
A declared process: what should happen, who owns it and the result to expect. Actual behaviour is compared against that.
- 03
How does it work with the other Crews?
It watches the boundaries between them, so a handoff from one Crew to another is watched.
Know what’s working.Catch what isn’t.
Hand over the watching. Keep every repair decision.
Your context. Your rules. Your final say.
Run the company. Not the agents.
Crew One is the operating system for AI-native companies.