Skip to main content

Erik's 6 Decisions

Erik's response to the kickoff thread. These are the operating decisions for the project.

Erik's read on the kickoff

"Your starlight-md-prototype.netlify.app with the deck all in one place is super compelling. I like the 3 big bets/moat — and what is common here is usability innovation. Your 7 demo moments up front is top tier product work. I appreciate the opinionated 7 sprints. Most likely the epics you have identified will go FSTR than you have projected — because they are well defined and we have clickable screens. We need to explicitly create engineering time for Security and Regulation Compliance."

1. Approve the scope

Approved. The scope as defined in the Brief, the Engineering Plan, and the prototype.

2. Pod size

We have a customer zero (Dr. P) and strong conviction. Erik wants the most development bandwidth practical.

Core pod (full-time interest already)

PersonRole
VictorEngineering
IlliaEngineering / brand polish co-lead
JulieEngineering
KevinEngineering

These four already have personal interest in Starlight.

Fractional support

Person/roleCoverage
Poy / AllanSecurity and infrastructure (fractional)
QAA #1 (and ideally #2)Quality assurance (fractional)
Why two QAAs

Two QAAs gives us parallel coverage on the clinical-quality and Atlas-parity tracks without making either a bottleneck. One is the floor; two is the target.

3. Tech stack as per Nao, Victor, Ken

Approved. See the full decision and rationale on the Tech stack page. In short: keep what's already working in the Starlight repo (Postgres + Drizzle, AWS via SST, Next 16, Cognito, MUI), add Stripe + Twilio + Claude/Whisper + Expo first, layer DoseSpot + LiveKit + Health Gorilla + Vanta as the roadmap demands, don't replatform.

4. Pilot terms

"I am of two minds. (1) Of course I am personally totally cool to build and see if Dr. Y finds it useful, her feedback alone is worth more than the price of the revenue. However, in order to BE SURE that we are creating something actually valuable it would be great for Dr. Y to start paying $ at the target price we would charge others when she is 100% confident that it is ROI+ for her practice."

The two minds

PathArgument
Free pilot for Dr. PHer feedback alone exceeds the value of any revenue we'd collect. Lets us iterate fast without commercial pressure.
Paid at target price once ROI+ confirmedForces us to be sure the product is genuinely valuable. Avoids "she's being polite" risk. Establishes the price discipline before sales.

The synthesis

Free during build-out and ramp. Move to paid at target price the moment Dr. P herself says she's 100% confident the product is ROI+ for her practice. Her conviction is the trigger, not a calendar date.

This both honors Dr. P's contribution (she gets the build for free) and disciplines us (we don't ship something she wouldn't pay for).

5. GitHub handles

"From the polaris repo you can find all of our github handles..."

TODO: Pull the team's GitHub handles from the polaris repo and:

  1. Invite all to dsurana04/starlight-md
  2. Add to a Starlight team in the org
  3. Set CODEOWNERS for the relevant pod

Erik already invited; Victor — DM Deepak your GitHub handle.

6. Brand polish

Lead by Deepak Surana & Illia.

The visual quality bar is set by the prototype. Brand polish — design system, marketing surfaces, prototype fidelity in the production app — rolls under DS + Illia.

What's next

StatusItem
In motionDr. P reviewing the prototype for clinical feedback
PendingPod members confirm allocation to Starlight
PendingFractional QAA(s) sourced
PendingSecurity/compliance work line-itemed into the sprint plan
PendingGitHub handles pulled from polaris repo, repo access granted