Black Flag Design
Concept · For Carey · WellSelf 360 Open the canvas →
A note to Carey · May 2026

From synthesis
on paper, to
synthesis in product.

We read everything you sent. The market position, the process flow, the framework logic explainer. Three documents that already do the hardest part of the work — the synthesis exists, the evidence is real, and the worked example holds up under inspection. What's missing is the shape Jordan and her care team will actually meet. This page is our opinion on that shape, and an invitation to build it with you.

— BFD concept v1 · for Monday's call
The bet we'd make
A synthesis people can hold in their hands beats a synthesis that lives only in a PDF. Build the product, the partnerships follow.
BFD → Carey↗ Phase 0 in 2 weeks
01 / What we read

Your work, as we received it.

These three documents are the foundation. They're rigorous, they're internally consistent, and they answer different audiences. We've rendered them here so the rest of this page can build on them, not restate them. Click any card to open it inline.

02 / What we noticed

Four things that change the build.

Read with a designer's eye, not a clinician's. These observations shape what we'd build first and why we'd build it that way.

01
The intake is the product — not the wrapper around the product.

Twelve sections, three layers, 28–35 minutes. Jordan is doing real synthesis work just by completing it. The intake itself needs to feel like care, not a form. Health-literacy calibration, recovery saves, breath room between sections — that's design, not copy. We'd treat the intake as the first surface that earns trust.

02
The Health Story has to read like it was written for one person.

The framework stays in the background; the lead drivers come forward. Jordan's version reads differently from Maya's because their physiology is different — not because we changed a template variable. Progressive disclosure is the whole design problem here. Get this right and the document survives being printed, forwarded, and screenshotted into a group chat.

03
The Care Team Companion is the trojan horse for clinician pull-through.

Wellframe, Twill, Welldoc — every patient-self-management tool has died on the same hill: clinicians don't read the patient-facing thing. The Clinical Edition has to land in under sixty seconds. A psychiatrist reading Jordan's profile should see the inflammation/CEP story without reading the educator's preamble. Drillable detail, library links, one screen.

04
Editorial review is the unit economics — design has to protect it.

Every output gets a certified educator's pass before it ships. That's the moat and the throughput bottleneck. We'd design the review surface as a first-class screen: a side-by-side where you can adjust depth, swap framing, accept/reject driver weight, leave a footnote — without ever leaving the document. Faster editorial = more Stories per week = the business works.

03 / The synthesis, live

What it looks like when it's running.

A small live sketch using Jordan's intake data. Toggle the symptom items on the left; watch the driver profile recompute; read the lead-driver call that would anchor her Health Story. This is the engine you described, with an interface around it.

1 Intake signal — section 3 + 4
2 Driver profile — recomputing
Driver weights are computed from the intake signal × condition library activations (here: MDD + Hashimoto's + fibromyalgia). The two highest become "lead drivers" — the synthesis the Health Story is built around.
Jordan, 44
My Health Story

Toggle symptom items on the left to see the synthesis form. Try unchecking everything to see how the story disappears — the engine doesn't fabricate.

Read the full Story in the canvas
Live computation against the same library entries cited in Stage 2 of the Framework Logic Explainer. click any intake item to toggle
04 / The conversation, playable

Play it through, once.

A live walkthrough of the design direction we'd build first. Click the suggested replies on the bottom — Lena (the educator) responds, the synthesis appears, the Action Plan forms, and the Care Team Companion ships at the end. This isn't a video. The state on the right updates as you go.

J
Jordan, 44
session 2 · with Lena Maren, NBC-HWC
Intake Synthesis Action Plan Care Team
↻ restart
Pick a reply to advance ↓
Where we are
Intake
Synthesis
Action Plan
Care Team Companion
Driver picture · live
Drivers light up after we hand off to the synthesis engine in §03. Right now we're still listening.
In her own words
"fatigue that doesn't lift with sleep"
05 / Three directions

Same synthesis. Three vocabularies.

We explored three end-to-end design directions on the canvas. Each one carries the full happy path: intake, Health Story, Action Plan, Care Team Companion. The card we've called out is the one we'd build first — and why.

Why we'd start with the conversation. It's the direction that scales editorial review (every accepted message is a draft of the Story), gives the Care Team Companion a natural transcript to drill into, and produces the lowest cognitive load on first contact. The other two stay on the table — Direction 01 becomes the downloadable artifact, Direction 03 becomes the Clinical Edition's at-a-glance view. They're not competing — they're a system.

06 / Phase 0

Two weeks to synthesis in product.

A concrete proposal. Two weeks, two people from BFD, working directly with you and one to two of the feedback testers. End state: a clickable prototype that produces a real Health Story for a real intake, with editorial review built in.

Week 01 — Intake + synthesis surface
  • D1–2Engine adapter. Wire Carey's library entries + Jordan's intake into a synthesis stub. Real driver weights, real mechanisms, no fabrication.
  • D3–4Intake, 3 sections deep. Section 3 (body), 4 (clusters), 11 (tailoring). Health-literacy calibration live.
  • D5Internal review. You + one tester walk it end to end. We rewrite copy where it's wrong.
Week 02 — Story + editorial review
  • D6–7Health Story v1, the conversation direction. Lead-driver anchor, progressive disclosure, save + share.
  • D8–9Editorial review surface. Side-by-side. Adjust depth, swap framing, accept/reject driver weight, leave a footnote.
  • D10Real Story, real tester. One Phase-0 tester completes intake. You review. We watch. We ship the prototype with their permission.