🎥 Telehealth (LiveKit)

1080p video on weak Wi-Fi. No patient downloads. Picture-in-picture so the clinician can chart while talking. Built on LiveKit’s selective-forwarding-unit (SFU) infrastructure — Synalux runs the rooms, you run the care.

🔌 Bandwidth-Adaptive Video (LiveKit SFU)

Synalux uses LiveKit’s open-source SFU rather than an MCU mesh, so:

• Per-participant bitrate adaptation — each viewer receives a stream sized to their downlink. Patient on hotel Wi-Fi sees 480p; clinician on fiber sees 1080p simulcast.
• No transcoding load on the server — CPU stays low even with 8-person multi-clinician case conferences.
• TURN fallback — if direct P2P fails (NAT / corporate firewall), the SFU relays via TURN.
• Track-level pause — bandwidth saver pauses outbound video when minimized to PiP.

🩺 Clinician-First UX

• No patient downloads — runs in-browser via WebRTC. Patient gets a one-tap join link via the Mail or SMS module.
• Picture-in-Picture (PiP) — minimize video, keep charting. Drag PiP anywhere on screen.
• In-call clinical notes — live SOAP draft via Voice Dictation running alongside the video stream.
• Patient chat sidebar — text chat in-call for quiet check-ins (lab values, dosing) without interrupting speech.

🎬 Recording (consented, encrypted)

• Per-call consent gate — visible on-screen confirmation before recording starts.
• Storage — recordings written to the Drive module with the same RLS isolation as other clinical files.
• Retention — workspace-configurable; default 30 days.

🔄 Conferencing Routing (in-event creator)

The Calendar module creator picks the right conferencing provider based on tier + workspace defaults:

🏗️ Architecture

POST /api/v1/livekit/token         Mint a per-participant LiveKit JWT (workspace + role-scoped)
POST /api/v1/meetings              Create a telehealth meeting (resolves room URL + token)
GET  /api/v1/meetings/:id          Read meeting status + participant list
PATCH /api/v1/meetings/:id         Update (extend duration, add participant)
DELETE /api/v1/meetings/:id        End meeting + flush recording to Drive
POST /api/v1/meetings/:id/invite   Send join link via Mail / SMS
POST /api/v1/meetings/:id/rsvp     Patient RSVP confirmation
GET  /api/v1/meetings/verify-token Validate meeting join token
POST /api/v1/calls/generate-token  Generate per-call auth token
GET  /api/v1/video/auth-token      Video auth for recording
GET  /api/v1/video/ice-servers     STUN/TURN server list
POST /api/v1/video/consent-log     Record consent for recording

⚖️ HIPAA + Audit

• Per-room access logging — every join writes to livekit_access_log with participant + role + timestamp + IP.
• TURN traffic encrypted — DTLS-SRTP end to end; no clear-text media on the wire.
• Recording consent enforced in code — backend rejects egress requests without a stored consent receipt.
• No third-party analytics in the call — LiveKit pages are CSP-locked.

💳 Plans

See full pricing → 

🔄 Inter-Module Integration

• Calendar — meetings auto-attached when an event is created with conferencing.
• Patients — every meeting is filed against the patient chart with consent receipt.
• SOAP / Clinical Notes — in-call dictation produces a draft note attached to the meeting record.
• Drive — recordings land in patient-scoped folders.
• Mail / SMS — join links sent via the connected channel.