AI workflow management · post-stabilisation ownership
No recurring fee without a recurring responsibility.
After accepted launch and 30-day stabilisation, your firm decides who monitors the system, coordinates providers, handles exceptions and governs change. Handover is a complete option — not a disguised path into a retainer.
Human-reviewed workflow
WORKFLOW / CONTROLLEDEvery transition keeps an owner.
01CaptureContext + owner
02RetrieveApproved sources
03PrepareLinked working material
04ReviewNamed authority
05RecordAccepted handoff
CONTROLLED CENTREAI prepares. A named person decides.
01 / The custody question
Who keeps client delivery working after launch?
Client-delivery workflows need an owner after launch. Sources, methods, templates and providers change; exceptions need decisions. Agree who handles that work and which checks are included in the operating scope.
Included in installation
Use-driven repair
Included in every installation. A deliverable cannot be accepted while it rests on a claim nobody has re-checked — so any claim the work touches is repaired as a condition of using it.
Assessed at Blueprint
Scheduled assurance
For firms whose facts change faster than their engagements use them. A recurring review reaches the claims that active work never touches.
Assessed at Blueprint
Carried operation
For firms with no internal owner who has real capacity, fast-changing sources, or a heavy exception load. The custody role is performed rather than handed over.
How the answer is reached: the Blueprint measures how fast the firm's sources actually change, how often client engagements run, and whether a named internal owner has real capacity. Where active use repairs claims before they are needed, scheduled or carried custody is not warranted and is not recommended.
02 / Commercial form
Each model is one custody answer, priced.
These are not service tiers with escalating value. Each is the commercial form of a single custody answer above, and the correct one follows from the operating condition rather than from preference. Prices shown are the current model, not an automatic continuation; final responsibility, hours, usage and dependencies are defined in the governing agreement.
Client-Owned Handover
No recurring Build n Bloom feeCarries · Use-driven repair
Warranted when
Warranted when you have a named internal owner with real capacity, your sources change slowly, and the work recurs often enough that active use repairs most claims before they are needed.
Build n Bloom owns
Build n Bloom owns handing over the accepted configuration, operating documentation, access map, failure path and handover record.
Client owns
You own monitoring, provider accounts, user administration, routine corrections and future changes.
Quarterly Assurance
Current model · US$2,500 per quarterCarries · Scheduled assurance
Warranted when
Warranted when you have an internal owner, but your facts change faster than your engagements consume them — so a scheduled review reaches the claims active use never touches.
Build n Bloom owns
Build n Bloom owns the named quarterly assurance work, within the agreed responsibility and hour boundary.
Client owns
You own daily operation, and escalating material changes between reviews.
Managed Optimisation
Current model · US$3,000 per monthCarries · Carried operation
Warranted when
Warranted when there is no accountable internal owner, sources change quickly, or the exception load is heavy enough that the custody role must be performed rather than handed over.
Build n Bloom owns
Build n Bloom owns the named recurring work within the agreed boundary; expanded implementation and usage remain separate.
Client owns
You own professional authority, lawful data, timely reviewers, and everything outside the contracted operating role.
03 / Responsibility map
The recurring work must be named before it is priced.
A broad promise to “keep the system working” hides both the risk and the labour. The operating agreement names the activity, the response boundary and the human authority that stays with your firm.
CONTROL / 01
Monitoring
Who notices failed runs, degraded retrieval, provider changes and unusual exception volume?
CONTROL / 02
Access
Who owns provider accounts, permissions, user administration and credential rotation?
CONTROL / 03
Exceptions
Who investigates an operating exception — and where does professional authority remain?
CONTROL / 04
Change
What counts as a correction, what counts as optimisation, and what requires a new scope or a risk review?
04 / Change control
Correction and expansion are different jobs.
The system stays governable when operating defects, routine assurance, optimisation and new scope are treated as different kinds of change.
01
In-scope correction
An accepted behaviour is not performing as agreed. During stabilisation, we correct and retest the in-scope defect.
02
Operating assurance
The accepted workflow gets a named inspection of its controls, versions, providers, exceptions and performance.
03
Optimisation
The same bounded workflow is improved, within an agreed responsibility and capacity envelope.
04
New scope
A new workflow, source, permission, integration, decision or risk boundary needs explicit assessment and agreement.
05 / Operating evidence
Keep value and reliability visible after launch.
The relevant measures are chosen during the Blueprint. They let the owner see whether the workflow is still earning its keep — without inventing a universal ROI guarantee.
Post-launch control loopAn accepted baseline can be observed without being silently rewritten
Monitoring creates evidence for a decision. It does not grant the system authority to change itself.
01
Run
Execute the accepted workflow version
02
Trace
Retain inputs, steps, latency and exceptions
03
Evaluate
Compare behaviour with written tests
04
Dispose
A person classifies defect, drift or new need
05
Correct
Change the bounded logic or control
06
Release
Retest and approve a new version
Normal operationAccepted version remains unchanged
Usage, retrieval, exceptions and reviewer behaviour remain visible to the named owner.
Material varianceStop, classify and choose the right change path
Defect, provider drift, new source or expanded scope follows its own authority and acceptance route.
01Retrieval
Time to find the applicable source, precedent or prior decision.
02Preparation
Human handling time from accepted intake to review-ready material.
03Review
Correction loops, unresolved questions and senior reviewer handling time.
04Evidence
Source coverage, required-field completeness and unsupported-claim stops.
05Handoff
Unowned commitments, missing decisions and time to an accepted operating record.
06Reliability
Exception routing, defects, rollback tests and version integrity.
07Claim currency
The share of material claims currently approved for their intended use — and the share past their review condition.
08Custody response
Time from an invalidating event to a recorded owner decision.
09Claim disposition
Claims re-approved, replaced or retired, rather than left unresolved.
10Reuse
Approved findings carried into the next engagement instead of rebuilt — the direct measure of whether the work compounds.
THE NEXT RESPONSIBLE STEP
Design the ownership model before the system is left to run.
The correct operating model depends on the workflow, internal owner, provider risk and recurring responsibility—not a preference for recurring revenue.
