Use these checks before you evaluate a case:
When a vendor says a system is live, it can mean a campaign, trial, phased rollout, or production. These are different levels of operational use.
When reading any case, use three checks:
- What public source supports the claim?
- What is the operating scope?
- What is clearly private?
1. Check the public source
Not every case has the same depth of published evidence.
External campaign source
Campaign material, event coverage, and other public content that supports a generative AI use case.
Examples on our site: Vinasoy and Enfagrow.
What you can trust:
- customer name
- campaign framing
- top-line volume or reach when publicly shown
What may still be private:
- infrastructure detail
- workflow design
- internal operating metrics
Customer-approved deployment material
Work has moved into customer-facing delivery, but no full public technical write-up is published yet.
Example on our site: GMA.
What you can use:
- the workflow and deployment pattern
- the customer-approved public wording
- the fact that the work moved beyond concept or demo
What may still be private:
- usage metrics
- internal screenshots and release records
- rollout depth by team or phase
- service-level architecture
Anonymized customer case
Real work where customer identity or operating details cannot be fully disclosed.
Examples on our site: APAC aviation and financial-services hyper-personalization.
What you can trust:
- the work is real
- the operating scope is described conservatively
What may still be private:
- customer name
- architecture
- performance detail
2. Check the operating scope
Deployment labels describe how far the work is integrated.
Campaign delivered
The campaign shipped and ran. This is common for consumer activations and personalized media experiences.
Trial deployment
The system is operating in a real customer setting at limited scale: more than a demo, not yet broad production.
Phase one rollout
Phase one is live. It does not mean the full roadmap is complete.
Production in customer environment
The workflow is running for real use inside the customer environment.
3. Check what stays private
Some details are public by design; others are not.
That usually includes:
- detailed architecture
- KPI and adoption detail
- internal rollout records
- infrastructure and IAM controls
That is normal in enterprise work. The key is clear boundaries between public and private detail.
A quick reader checklist
- What public source supports this claim?
- Is this a campaign, trial, phased rollout, or production system?
- Does the case state where the system runs?
- What details are explicitly kept private?