Skip to main content

Product Horizon

Post-v1 features and architectural directions for Starlight Practice itself — distinct from Auxiliary Bets, which describe entirely separate products.

Important framing: none of these are in v1 scope. v1 is the EMR + parent app + practice management + AI scribe captured in the Launch Briefing Prototype. The pages in this section describe where the product is heading after v1 is healthy and operational with Dr. P — they're written down so the team can pattern-match opportunities and so v1 architectural decisions don't accidentally foreclose them.

What's in this section

Two clusters

ClusterPagesTheme
AI in the parent flowParent Triage, Agentic Safety MonitorTake the parent app from passive (notifications, summaries, refill requests) to active — a 24/7 reasoning surface that's anchored to the patient's actual chart, with a safety net watching every conversation.
The end-state interfaceVoice-Driven Dynamic UXLong-arc architectural direction: voice as primary input, UI that hydrates itself in real time per task and context. Same data + APIs, regenerated presentation.

Gating triggers (when each becomes interesting)

PageBecomes interesting when…
Parent Triagev1 parent app has families using it, doctors are seeing real after-hours-message volume, and the Compliance · Synthetic Data Program is mature enough to safely test medical-reasoning loops.
Agentic Safety MonitorParent triage is shipping or about to. The two are designed together.
Voice-Driven Dynamic UXContinuous — the architectural decisions in v1 should keep the door open even if no voice-UX feature ships in v1 itself.

Cross-references