Services
Remote / Written

See it clearly, then know what to do.

This work sits on the user side: reading how trust gets allocated, how judgment changes, and how to see it before it happens. Tool rollout and productivity belong to tool-side consultants.

01
Index

Start from your situation

A conversation next week that cannot go wrong
How much judgment have I handed to AI
The product is about to ship
Everyone uses AI and the organization has no criteria
Locate the organization's risk hot spots
03
Standards

Shared delivery standards

Form
Everything is delivered in writing.
Scale
A four-level risk scale: low, medium, high, critical. Comparable across services.
Priorities
Remediation priorities written in full: must fix, improve within 30 days, next release.
Discipline
Readings state their evidence. Grades map to signals. Deliverables never judge personality.
Timing
Delivered in writing within the agreed window. Site-wide reply commitment: within one week by email.
04
Choosing

Not sure where to start

Fill in the screening form. If you are unsure, choose “not sure yet”. The reply will recommend a starting service, or say plainly that this is not a fit.

05
Boundaries

Boundaries and referral

Boundary

Not psychotherapy

No counseling, psychotherapy or clinical assessment.

Boundary

Not legal advice

No legal opinions or compliance determinations.

Boundary

Not security audit

No security testing or data protection audits.

Boundary

Not medical assessment

No medical or health assessment of any kind.

Boundary

Not engineering onboarding

No deployment, integration or architecture assessment.

“Risk can be identified and managed. It cannot be guaranteed away.”

If you or someone near you is in crisis, in Taiwan call 1925 (mental health support line), 1995 (Lifeline), or 119. Crisis handling always takes priority over any delivery schedule.

06
Data

Data handling summary

Material
Only de-identified material is accepted, and it is re-checked on receipt.
Use
Material is used for this engagement only. It is not used to train AI models.
Deletion
Source material is deleted within 30 days of delivery. Screening questionnaires within 90 days.
Quotation
Anonymous quotation requires written consent case by case. No selection means no consent.
Tools
When AI tools assist the analysis, this is disclosed and agreed in advance.

Full statement: Privacy and data handling.

Begin

Confirm fit first, then begin.

Fill in the screening form