What to inventory
Begin with AI-enabled processes used in customer service, sales, reporting, operations or internal decision support. For each one, record its business purpose, operator, owner, vendors, model dependencies, data inputs, system connections and fallback procedure.
Distinguish a production workflow from a pilot or occasional employee use. Ask which processes are material to revenue, service quality or continuity, and which can be paused without significant disruption.
Seven diligence questions
1. Who owns the workflow?
Identify the accountable business owner and the person who maintains its configuration. Are these responsibilities written down, and can they be transferred?
2. Who controls access?
Review administrator rights, vendor accounts, authentication, service credentials and permissions. Avoid assuming that access held by the founder or developer will transfer automatically.
3. What data and systems does it depend on?
Map source systems, permissions, integrations, retention requirements and relevant contractual restrictions. Verify that the intended operator can continue to use the necessary inputs lawfully and reliably.
4. How are outputs checked?
Determine where human review is required, how errors are identified and how the business handles an incorrect or incomplete result.
5. What happens when a dependency fails?
Ask how the process responds to a model change, vendor outage, expired credential, missing data or unavailable operator. Is there a tested manual fallback?
6. Can another person run and maintain it?
Request evidence that an authorized backup can execute the workflow, investigate an exception and restore service without the original builder.
7. How is performance monitored?
Identify the measures used to track output quality, cost, timeliness, failures and changes in business outcomes. A claimed productivity gain should be distinguished from a measured one.
Evidence levels and limits
An assertion that an automation is “fully documented” is not the same as a successful handoff. A procedure is useful evidence. Stronger evidence includes a documented test in which a second operator runs the workflow, resolves a problem and records the result.
Not every AI workflow warrants the same diligence depth. Prioritize those connected to essential operations, material business claims or customer-facing decisions.
After the acquisition
Ownership transfer should include access reviews, named operating responsibilities, monitoring, vendor contacts, fallback procedures and a schedule for reassessment. The goal is to ensure that the business retains the capability, rather than only the software artifact.
GrowFast's VGS methodology distinguishes between asserted, documented and operationalized evidence. Day Two then considers whether value continues to hold after the ownership change.
Related questions
Operational continuity diligence places AI within the wider operating environment. Founder dependency during a business sale addresses concentration of knowledge and decision-making beyond technology.