Journeys & AI with young people
Synced from Asana (20 Jul 2026): Journey Maps Document + AI Specialist interaction model with young people. Door-opener view while design continues.
Design intent for demos
These are the journeys and interaction rules the funded product must honour. The prototype shows the Need board and specialist catalogue; it does not yet run live AI chat with young people. Use this page to explain how AI is allowed to behave.
Core experience (young person + AI)
“I was listened to, I understood what was happening, I remained involved in the decisions, I could speak to a person when I needed to, and the support helped fulfil my Need.”
“I had to convince a machine that I deserved help.”
Every AI-supported experience must be: safe, respectful, honest, accessible, age appropriate, non-judgemental, easy to understand, easy to leave, connected to human support, and focused on fulfilling the Need.
AI interaction rules with young people
Never pretend to be a person or use a fake human identity to create false trust.
Young people can request a human at any stage without being disadvantaged.
AI must not finally decide safeguarding, significant eligibility, professional specialist assignment, outcome achievement, complaints, appeals or Need closure.
AI may organise or summarise — original words and meaning must stay available.
Before personal data: why needed, how used, who sees it, whether AI processes it, what if they refuse.
No emotional dependence, guilt, artificial urgency or excessive praise to influence choices.
Pause, save and return, correct information, decline where permitted, challenge recommendations, complain, request a person.
AI may surface risk indicators. Qualified people make safeguarding decisions. In crisis: acknowledge, stop routine questions, escalate.
Moments AI may support (examples)
- Explore before sharing personal information
- Ask for help / submit a Need
- Privacy, consent and confidentiality explained honestly
- Immediate-risk screening (brief, non-accusatory)
- Prepare for triage and assessment
- Describe strengths and aspirations
- Review intended outcomes and proposed matches
- Appointments, reminders, progress questions
- Evidence, outcome review, sustainment
- Correct information, challenge a recommendation, complain
- Request a human specialist
- Re-engage after missed contact
- Close or reopen a Need
Prohibited AI behaviours
- Pretending to be human
- Making final safeguarding or eligibility decisions
- Claiming “everything is completely confidential”
- Manipulating with guilt, fear or false urgency
- Blocking access to a person
- Discarding the young person’s own words
Confidentiality must be explained accurately — never “everything is completely confidential.”
Stakeholder journeys (Asana journey map framework)
Journey maps start from the stakeholder’s need, show real touchpoints and pain points, and must be validated with the people who live them. Future-state design builds on this spine.
From first contact to a sustained positive destination — with dignity and a real voice.
- Referral
- Registration
- Assessment
- Support planning
- Engagement
- Training & development
- Employer introduction
- Work experience
- Employment / education progression
- Ongoing support
- Programme completion
Simple, valuable hiring of ready candidates with support when things wobble.
- Initial contact
- Programme introduction
- Vacancy submission
- Candidate matching
- Recruitment
- Placement support
- Progress reviews
- Ongoing partnership
Hand off once; see the Need owned and progressed — not lost in email.
- Referral preparation
- Referral submission
- Eligibility confirmation
- Information sharing
- Progress updates
- Outcome reporting
- Ongoing partnership
One coherent board: ownership, risk, stalls, outcomes, learning.
- Case allocation
- Assessment
- Planning
- Delivery
- Monitoring
- Reporting
- Escalation
- Case closure
How this links to the prototype
- Needs board = where every stakeholder Need becomes visible.
- Specialists = capability directory including the 18 AI roles (recommend only).
- Processes = lifecycle, forms, RACI, general AI rules.
- Transition = Nemawashi, readiness areas, AI operating parameters before go-live.
- Demo cases (e.g. safeguarding NEED-2026-0004) show why human handover and risk rules matter.