ZenThink — Safety-Aware Conversational Wellness UX
Johnny's role: AI experience and safety architect—product intent, conversational workflows, safety boundaries, escalation requirements, privacy expectations, and human-centered experience design.
Safety-aware conversational wellness product exploring mood and journal workflows, context-aware AI interactions, escalation design, and non-diagnostic language — not therapy or medical treatment.
Recruiter summary
- What it is
- Conversational wellness product intent with mood/journal workflows and strict safety boundaries.
- What I directed
- Product intent, conversational workflow design, safety boundary requirements, escalation design, and human-centered experience patterns.
- What was hard
- Designing supportive AI interactions without crossing into diagnosis, treatment, or emergency-care claims.
- Evidence
- ZenThink repo in HQ workspace plus portfolio case study with COMING SOON status — no shipped public feature claims.
Product problem
Wellness AI must support reflection and journaling without implying diagnosis, treatment, or emergency care — especially for vulnerable users.
Product status
Coming soon — product intent and safety architecture documented; no public beta or shipped feature claims without safety review.
System designed
Conversational wellness UX with mood/journal workflows, context-aware AI interaction patterns, safety boundary layer, and crisis-resource escalation routing.
- →Conversational product design with mood and journal workflows
- →Context-aware AI interaction patterns
- →Safety boundary and escalation design
- →Privacy-conscious data handling patterns
- →Non-diagnostic language requirements
AI-assisted implementation
Defined AI-assisted conversation patterns and safety review criteria for future implementation — human-centered review before any public conversational features ship.
Safeguards and evaluation
- ✓Non-diagnostic language requirements for all AI responses
- ✓Crisis-resource escalation — not AI-only emergency support
- ✓Privacy-conscious data handling patterns in product intent
- ✓No fabricated wellness outcome or satisfaction metrics
Verified evidence
- ✓Portfolio case study with COMING SOON status label
- ✓Documented safety boundary and escalation architecture
- ✓ZenThink repo presence in HQ workspace (product maturity varies)
Known limitations
- •Not a licensed therapist or medical device
- •Not for emergency care or crisis intervention by AI alone
- •No public beta — features described are product intent
- •No verified clinical efficacy or anxiety-reduction claims
Technologies used
Next milestone
Complete safety review checklist and privacy governance before any public beta — crisis resources and not-medical-advice positioning at launch.
Related content
Interested in this kind of product work?
I'm open to roles where AI Product & Automation Systems Specialist skills apply — agentic systems, automation, and evidence-driven product delivery.