Lian Passmore Β· GEN25
Assessor
Map.
A complete directory of the Project Rise research repository.
project-rise / evidence-directory
Learning Outcome 1
Systematically plan, execute, complete and evaluate a work-based research project resulting in a substantial and original contribution to knowledge as evidenced by significant benefits for the profession.
Your agreement: Design, test & evaluate conversational AI prototypes culminating in Ray β contributing practical frameworks for culturally governed ethical AI.
Learning Outcome 2
Through engagement with a significant work-based project and reflective practice extend knowledge and skills developing personal potential and professional competencies in order to be effective in changing complex work environments.
Your agreement: Evolved from conceptual design to applied AI development β building conversational agents, integrating voice and text, applying kaupapa MΔori to ethical design.
The Journey Overview
Specific academic criteria mapped to journey sections.
Introduction Β· Video 2β3 min
Introduction
Positionality, research context, personal motivation
LO1
Research scope established. What problem, why it matters, why now.
Definitions & context Β· Text + hover glossary
GO: Technological Futures
10 MΔori and Pasifika concepts β the analytical framework, not a layer on top of it.
Research question Β· Visual (Kei Compass)
LO1
Main research question + five Kei Compass sub-questions. Each direction maps to an artefact.
Research Questions
Adapted from Dell (2025). Application to conversational AI design is original.
Methodology Β· Audio 4β5 min + timeline
Methodology
PAR + autoethnography + kaupapa MΔori. Four builds, ethics, consent, data analysis.
View Appendix C1: Ethics Official Ethics Form (MTF.8888.275) (Assessor View)LO1 β Systematic planning
Four builds in deliberate vulnerability progression. Each build informed the next β causality is the methodology.
LO2 β Reflective practice
Six named deviations G1βG6, each explained and justified. Full governance timeline in Appendix C2.
View Appendix C2: DeviationsGO: Leadership
Kaupapa MΔori governance of research process. Cultural supervisors: Lee Palamo, Nadine Young, Rob Ngan-Woo.
Findings preview Β· Artefact card
LO1 β Original contribution
Safety method across four builds. Answers Kei roto (Agency) + Kei mua (Values). Full artefact β
Primary case study Β· Card + live demo
LO1 β Evaluated prototype
30/70 finding, State Before Story, crisis architecture. Four builds led here.
LO2 β Technical capability
Live demo = voice-first AI built and deployed. Working prototype is LO2 evidence.
GO: Technological Futures
Applied AI development at the frontier of ethical design.
Findings preview Β· Artefact card
LO1 β Original theory
Human Proxy Theory + Voice as Justice. First application of Teu le vΔ to AI design.
Central finding Β· Quote + silence
LO1 β Unresolved paradox
Tapu/noa paradox. Held without resolution β that restraint is the scholarly position. Full evidence in Relational Space.
Findings preview Β· Artefact card
LO1 β Original framework
DreamStorm Charter. Benefit to profession: reusable by any practitioner.
LO2 β Values under pressure
Build Code discovered under constraint across four builds β not planned in advance.
GO: Leadership
Ethical development as a leadership practice, not a compliance task.
Discussion Β· Text ~2000 words
Discussion
25 citations. Five threads mapped to five Kei Compass directions.
LO1 β Theories in context
Human Proxy Theory (Goffman challenged). Equity-Safety Paradox (Noble/Gebru + compute cost). Values as architecture (Manahau). Voice as justice (Ong).
Critical reflection Β· Audio/video + LO table
LO1 β Full evidence hub
Audio traces research journey. Table maps each artefact to the LO1 contribution it evidences.
LO2 β Full evidence hub
18-month competency development. Who I was vs who I am now. Dated study group recordings are the spine.
View Appendix C3: Study Group EvidenceGO β All three themes
Technological Futures: AI design capability. Leadership: kaupapa MΔori governance. Value to Communities: utu tΕ«turu, knowledge return.
Learning Agreement
LO1 statement: design, test, evaluate Ray. LO2 statement: competency from conceptual to applied AI development. Both fulfilled.
Official Learning Agreement (Assessor View)AI reflection (required) Β· Text ~2000 words
AI Reflection
Why, how, where, what I learned. Claude, Wispr Flow, NotebookLM, ElevenLabs, Perplexity. Prompt evolution: transactional β systemic.
View Appendix B3: Prompt EvolutionLO2 β Professional judgment
Voice Guardian skill built in Claude = technical solution to professional problem. Five-round drafting process.
Conclusion Β· Text ~300 words
Conclusion
Three key findings in three sentences. Open questions. Implications for practice.
LO1 β Benefits to profession
Sovereign infrastructure. Community governance. Three years of relational groundwork.
GO: Value to Communities
Utu tΕ«turu β the loop must close. Path forward named honestly.
The Artefacts Detail
standalone documents containing the deep methodology, practice, and code evidence.
Primary Case Study
Ray: AI Relationship CoachLO1 β Values-to-architecture
Six MΔori values each translated into a hard-coded architectural rule. Original contribution: values as structure.
Appendix A4LO1 β Evaluated prototype
59 pilot sessions. 30/70 finding, State Before Story, crisis flag on session one.
Appendix C4LO2 β Ethical judgment
Crisis architecture built after session one revealed the risk. Real-time decision = LO2 professional competency.
Appendix B2LO2 β Technical capability
Live working prototype: Next.js, ElevenLabs, Supabase, Vercel, Claude. Deployed. Interactive.
Appendix B4Method Artefact
Building Safe Conversational AILO1 β Original practitioner framework
Nine builder principles + vulnerability ladder. Not adapted from literature β emerged from four real builds.
Appendix A2LO1 β Systematic execution
Four builds documented: what was tried, what failed, what changed, and why.
LO2 β Deviations G1βG6
Six named methodological deviations. Each explained and governance-approved.
Appendix C2LO2 β Learning through building
The building was the thinking. Each build informed the next. Reflective practice in the research itself.
Appendix B2Thought Piece
Conversational AI as Relational SpaceLO1 β Human Proxy Theory
Original: AI borrows trust from human accountability structures. Goffman (1959) challenged. Empirically grounded in W04.
LO1 β Tapu/noa paradox
Central unresolved finding. Held without resolution β that restraint is the scholarly position. Marsden, Taiuru, Hudson engaged.
LO1 β Voice as justice
First application of Teu le vΔ to structural AI design. Oral tradition literature connected (Ong, 1982).
GO: Value to Communities
VΔ as a returnable design framework. Knowledge goes back to communities, not just about them.
Framework Artefact
Build Code PracticeLO1 β DreamStorm Charter
Original practitioner charter: reusable by any builder. Benefit to profession.
Appendix B1LO2 β Values under pressure
Wall of NO: every decision where a value held against pressure. Professional competency as practice.
LO2 β Competency progression
Build codes evolved: Project Rise (simple) β Ray (architecturally sophisticated). Development documented in real time.
GO: Leadership + Community
Open-source framework. Make your own. Returnable knowledge, not claimed ownership.
Collaborative Framework
Bonus: Tikanga FrameworkLO1 β Original synthesis
Merges relational governance (Ohu) with technical architecture (Mahi). Validated by Dr Karaitiana Taiuru.
LO2 β Professional practice
Late-stage cross-cohort collaboration resulting in a functional, open-source industry tool.
Academic Storytelling
Creative ArchivesLO2 β Creative methodology
Translating research into podcasts, digital storybooks, and visual media. Demonstrates range of communication competency.
GO: Value to Communities
Research made accessible beyond academic audiences. Returnable knowledge in formats communities can actually use.
Creative Academic Practice
Presentation video + podcast + digital storybooks. Evidences the "evolving creative academic methodology" arc of the project.
Standalone Content
Required written elements for He Rourou and formal report documents.
Abstract
β250 words. States problem, method, findings, conclusion. Required written element for He Rourou.
Executive Summary
β750β1000 words. Maps LO1 (what you did) and LO2 (how you grew). Published in He Rourou proceedings.
References
β25 APA citations. 15β25 range. Engages MΔori, Pasifika, and Western scholarship.
Glossary
β43 terms. Powers hover tooltips site-wide. Full list with definitions.
Appendices Index
Raw data, technical logs, ethics forms, and decision traces.
LO1
The Kei Compass Framework
LO1
Vulnerability Progression Framework
LO1
Coding Frame & Thematic Categories
LO1
Build Code, NO Clauses & DreamStorm Charter
LO1 + LO2
Safety Decision Traces
LO2 (AI)
System Prompt Excerpts & Safety Architecture
LO2
Technical Architecture Overview
Struct
Ethics Documentation
LO2
Deviations & Governance Timeline
LO2
Study Group Evidence & Reflective Practice
LO1
Participation & Engagement Summary
LO1
Participant Questions by Build
Struct
Participant Communications (Sample)