# Asset needs — agetech

---

## Images

### 1. Service blueprint (Chapter 01)
- **What it is:** The V4 service blueprint (`Aging _ Blueprints Working - Aging, Service Blueprint, V4, 04_26.svg`) showing the full service arc from caregiver onboarding through forecast through ongoing guidance.
- **Where it likely exists:** `raw-case-material/agetech/Aging _ Blueprints Working - Aging, Service Blueprint, V4, 04_26.svg` — also `raw-case-material/agetech/service-blueprint.svg` and `raw-case-material/agetech/blueprint-blurred.png` (already blurred).
- **Why the rewrite needs it:** Chapter 01 establishes the scope of what Whitney built. A service blueprint at this slot anchors the "prototype had to actually behave like the service" claim visually before the reader commits to Chapters 02–05.
- **Fallback:** Use `blueprint-blurred.png` from raw-case-material if the SVG requires redaction work. Alternatively, a wide-format screenshot of the GPT architecture diagram (experiment planning area from `gpt model builing notes.pdf`) that shows the system's component structure.

### 2. Questionnaire tiers diagram (Chapter 02)
- **What it is:** A clean diagram of the three-tier question architecture (priority / secondary / context) with branching logic illustrated.
- **Where it likely exists:** `raw-case-material/agetech/determining wireframe components and uncertainties.pdf` or `wireframing and sketching.pdf` may contain a working version. The `QuestionnaireTiers` component in the live MDX was a custom Astro component — its source content is the closest existing rendering.
- **Why the rewrite needs it:** Chapter 02 describes the architecture decision to tier the questions; the diagram makes the structure legible at a glance and breaks up a text-heavy section.
- **Fallback:** A screenshot from the Spotlight deck (`04-09-2026 Spotlight - Custom GPTs for Service Design (1).pptx`) showing the questionnaire or question structure slide. Or a greybox placeholder labeled "questionnaire architecture — 3 tiers."

### 3. Interaction model comparison (Chapter 04)
- **What it is:** Side-by-side comparison of v1 (sequential pipeline, 6 phases, value at the end) and v2 (modal tool, 8 task-shaped modes, value immediate).
- **Where it likely exists:** The `InteractionModelComparison` component in the live MDX was a custom Astro component — its visual treatment is the closest existing reference. `gpt-v1-files/` and `gpt-v2-files/` in raw-case-material contain the actual prompt architectures. The Spotlight deck may have a slide showing the structural shift.
- **Why the rewrite needs it:** This is the case's pivot moment. The chapter says "v1 asked. v2 helped." The diagram has to show what that meant structurally — not just in words.
- **Fallback:** A screenshot of the `interviewboard.svg` or `interviewboard-blurred.png` from raw-case-material as the v1 evidence, paired with a `chatgotlofi.svg` screenshot as the v2 reference. Or a greybox labeled "v1 vs v2 interaction model comparison."

---

## Pull quotes

### Primary: "I just gave you all that information" (Chapter 03)
- **Attribution:** v1 participant — this is already in the live MDX. Confirm it's from `cohort-1-chat-transcripts.md` and that the anonymized attribution ("a cohort 1 participant") is sufficient.

### Secondary: The executive demo (Chapter 05)
- **The quote:** *"We can't get him to stop"* — attributed in the live MDX to "Note from the session."
- **Clarification needed:** Is this quotable? Should the attribution be "facilitator's session notes, March 2026" or similar? This is the case's strongest proof point.
- **Source:** `stakeholder-chat-transcripts.md` in raw-case-material — confirm the quote appears there in a quotable form.

### Optional: "This would have freaked us out at the time" (Chapter 05)
- **Attribution:** Executive stakeholder, during the post-study demo — confirm this is in `stakeholder-chat-transcripts.md` and that the anonymized attribution ("the executive sponsor") is sufficient per CASE-CONTEXT.md redaction rules.

---

## What Whitney needs to confirm

1. **Beta timing:** The scaling section says "funded for fall 2026." Confirm whether the beta has since run, and if so, replace "funded for fall 2026" with the actual outcome.
2. **"We can't get him to stop" quote:** See pull quotes above — attribution and quotability.
3. **Blueprint redaction:** Is `blueprint-blurred.png` cleared for publication? Or does the SVG version need additional redaction before use?
4. **Participant count footnote:** The case says "12 caregivers tested across two structural versions." The live MDX confirms this (7 on v1, 5 on v2). Confirm this is accurate and publishable.