No signed profile
Manage identity, current session, role boundaries, and audited activity for this staff account.
Sign in as a clinic staff member or VoiceOps super admin to view profile, role, and access boundaries.
Today-first operating view for appointments, missed work, escalations, and human-loop tasks.
Manage identity, current session, role boundaries, and audited activity for this staff account.
Sign in as a clinic staff member or VoiceOps super admin to view profile, role, and access boundaries.
Manage clinic tenants, launch readiness, provider connections, support access, and cross-clinic compliance.
Owner/admin control center for clinic identity, operating rules, automation safety, communication, and audit evidence.
Add at least one clinic location before live voice and WhatsApp routing.
Branch records are the source for appointment location, WhatsApp location links, report desk routing, and future per-location phone numbers.
Voice booking should only offer approved doctor slots and must escalate same-day exceptions to reception.
No settings audit event yet. Save settings or update staff to create audited evidence.
Use analytics as a pilot evidence board: find recovered work, unresolved human-loop risk, knowledge readiness, and operating areas that need staff review before clinic scale-up.
These are local operational metrics for pilot validation. Production ROI reports will compare recovered calls, bookings, no-shows, and staff time saved.
Use this view as the pilot evidence board: every number links to an operating workspace where staff can inspect the underlying calls, messages, reports, integrations, or escalations.
Analytics is read-only. It should guide staff review and owner decisions, not trigger medical, billing, or migration decisions automatically.
0 completed calls
2026-05-04 schedule
Ready for clinic flow
0 urgent follow-ups
0 registrations
0 delivery issues
Use these tabs to jump to the exact clinic work area. Each section below opens from its header and can be minimized after review.
No patient records are available yet.
Reception can scan the day, confirm visits, recover no-shows, and queue WhatsApp reminders from one workflow.
Review voice-agent outcomes, linked appointments, failed tool calls, low-confidence cases, and staff handoffs before production activation.
The voice agent can book appointments, check slots, find patients, and create staff escalations. It cannot diagnose, prescribe, interpret reports, or make emergency decisions.
Real Vapi/Bolna calls stay blocked until assistant binding, phone routing, public callback URL, webhook security, and internal test-call evidence are all ready.
Production mode uses live voice-provider webhooks and backend tool APIs. Mock voice configuration and simulated calls are disabled.
Track report-ready handoffs, secure-link delivery, failed patient notifications, active link expiry, and doctor-review escalations without allowing AI report interpretation.
Manage Native, Connect, and Migration modes with clear ownership rules, connector readiness, mapping review, and failed-sync recovery before data reaches voice, WhatsApp, reports, or analytics workflows.
Connect Mode reads and maps external objects. VoiceOps does not overwrite external CRM/HIS/LIMS data automatically; writeback belongs to a later approved migration step.
Production integrations accept real connector credentials, file imports, and provider webhooks only. Demo sync failure tools are disabled.
Use this as the service-provider checklist for the selected clinic. It separates local setup, staging proof, and live patient-traffic approval in plain language.
Map every live external value to the exact env variable, screen, API route, and evidence.
0 blocked item(s). Use this before deployment handoff.Prove production data, auth, deployment URL, backup, restore, and rollback behavior before provider activation.
0 blocked item(s). This is the rehearsal bridge before live provider traffic.Rehearse public URL, webhook routes, signatures, provider callback registration, staging env parity, and rollback.
0 blocked item(s). No real provider traffic is enabled here.Choose the staging host, prepare env templates, copy callback URLs, and document rollback before provider activation.
0 blocked item(s). This turns live setup into an operator checklist.Run the final local and staging-preview checks for env, admin, database, callbacks, rollback, and live-readiness gates.
0 blocked item(s). This is the last local pack before staging deployment values are entered.Move from local JSON prototype storage to the production database safely.
Check Prisma models, database URL, migration marker, and restore readiness.Store voice, telephony, and WhatsApp setup details without exposing secrets in the browser.
0 masked credential profile(s); callbacks and public URLs still decide live readiness.Test provider callback shapes with sanitized samples before real patient traffic.
0 fixture sample(s). This proves parser contracts only; it does not send calls or messages.Confirm production auth, persistence, provider, webhook, and rollback gates.
0 blocker(s). Localhost and mock evidence cannot unlock patient traffic.Run a no-live-patient-data rehearsal with synthetic calls, messages, reports, and rollback evidence.
0 synthetic evidence record(s) recorded for dry-run confidence.Final gate for real patient calls, WhatsApp messages, reports, and clinic data.
0 live evidence record(s), 0 signoff approval(s).Live patient traffic remains blocked until the deployment gate, Postgres gate, provider callbacks, staging rehearsal, live evidence, and owner/super-admin signoff all pass. This panel is operational guidance; it does not bypass backend safety checks.
This pack is safe to share with operators because it never returns raw API keys, tokens, passwords, patient names, phone numbers, transcripts, or report data.
One operator checklist for Postgres cutover, admin bootstrap, auth/session lockout, deployment URL rehearsal, and rollback proof.
The pack is operator-safe: it returns commands, env variable names, routes, artifacts, and evidence names only. It does not return database URLs, passwords, auth secrets, provider tokens, patient names, transcripts, or reports.
Rehearse the deployed URL, provider webhooks, callback signatures, staging env parity, and rollback path before real provider traffic.
This pack returns only route names, env variable names, artifact names, and evidence requirements. It does not return raw webhook secrets, provider tokens, connection strings, patient data, callback payloads, transcripts, or reports.
Choose the deployment target, prepare staging and production env templates, map callback URLs, and document rollback before real provider activation.
Staging and production values must stay separate. This screen can be shared because it returns placeholders, route names, env variable names, and evidence requirements only. It does not return API keys, passwords, tokens, connection strings, transcripts, reports, or patient records.
Prepare the final env checklist, admin bootstrap, database migration, callback map, preview smoke test, rollback, and live-readiness checks before staging deployment.
This rehearsal is local and staging-safe. It returns command names, env variable names, route names, and evidence requirements only. Real provider tokens, passwords, database URLs, patient data, transcripts, and reports are never returned.
Separate safe local setup from staging dry-runs and real live-provider evidence before any patient traffic.
Vapi primary with Bolna secondary fallback. Stored keys remain backend-only.
Default routing is direct Vapi/Bolna. Exotel SIP/DID is optional for clinics that need it.
WhatsApp Cloud/BSP credential and template/callback evidence are required before live messages.
Staging callback dry-run needs approved Vapi/Bolna voice, WhatsApp, and rejected-negative fixtures. Exotel is conditional.
Use this area to prepare blank provider wiring locally. Saving a credential stores an encrypted backend record only; verification is a separate deliberate backend action after you provide real keys.
Use this before external-provider activation so a non-technical service provider knows exactly what is missing, where to enter it, and what proof is required before live patient calls or WhatsApp messages.
Use one stable HTTPS domain for every provider dashboard. Localhost and non-HTTPS URLs stay blocked.
Set VOICEOPS_PUBLIC_APP_URL after deployment, then copy the route URLs below.HTTPS public URL readyVoice, Exotel, and WhatsApp callbacks must fail closed unless server-side secrets/tokens are configured.
Set VOICEOPS_WEBHOOK_SECRET plus WhatsApp app secret/verify token and optional Exotel token on the server.Server callback secrets readyDefault route is Vapi primary with Bolna fallback. Exotel/SIP is optional only for clinic-owned DID/SIP forwarding.
Store and verify the Vapi key first, add Bolna as secondary, then record a real internal test call.Voice credential and route readyManual WhatsApp can work locally, but automated reminders, registrations, reports, and feedback require Cloud/BSP sender setup.
Collect WABA, phone number ID, approved templates, app secret, verify token, and one consent-safe test message.WhatsApp credential and templates readyPatient traffic remains blocked until provider fixtures and real call/message/callback evidence are recorded.
Approve sandbox fixtures, run staging dry-run, then record real test call and message evidence in Live Readiness.Sandbox and live proof ready/api/voice/webhooks/vapiblocked/api/voice/webhooks/bolnablocked/api/clinic/notifications/webhooks/whatsapp-cloudblockedbooking_confirm, reminder_1h, report_ready, feedback_request, opt_outrequiredRefresh callback ops to load voice and WhatsApp callback metrics from the backend policy boundary.
Raw callback payloads, signatures, phone numbers, transcripts, message bodies, recording URLs, tokens, and secure links stay server-side.
Record or refresh provider fixtures to review real callback contract readiness before any live patient traffic.
Fixture review stores hashes and masked references only; raw provider payloads and headers stay out of browser state.
Production launch needs real staff login, Postgres persistence, demo session disabled, Vapi/Bolna voice credentials, selected call-routing proof, WhatsApp credentials, webhook signatures, encryption, and audit review before live patient calls.
Track each real provider value and test result here after deployment and paid-provider access are ready. This is the live evidence ledger; staging fixtures and demo data do not unlock patient traffic.
Final live patient data signoff requires fresh backend provider verification, live voice/routing/WhatsApp round-trip evidence, clinic owner/admin approval, and VoiceOps super-admin approval. Mock or localhost evidence is rejected in production.
Phase 11M is a rehearsal packet only. Keep all provider tests on internal numbers and synthetic patients until Live Readiness and the activation packet both pass.
Phase 11N is a dry-run ledger. It records synthetic evidence for staging HTTPS, auth, Postgres restore, callback fixtures, patient workflows, monitoring, rollback, and clinic go/no-go without mutating Live Readiness or calling paid providers.
These are local verification commands only. They do not call Vapi, Bolna, Exotel, WhatsApp, or any paid provider.
npm run db:local:upStarts the local database container used for cutover rehearsal.npm run db:migrate:deployApplies Prisma migrations before the app trusts Postgres tables.npm run db:verify:localChecks DB connectivity, migration table, and key PatientOps record counts.npm run db:bootstrap:super-adminCreates the first service-provider admin only when the staff table is empty.npm run db:cutover:dry-runRuns production-like auth, Postgres persistence, browser validation, and screenshot proof locally.Cutover approval should be based on restored backup counts, successful custom-login proof, seeded super-admin evidence, demo controls disabled, and a clean browser run. Keep clinic patient traffic blocked until Live Readiness is separately approved.
Open service-provider access only with owner approval, verify every support event in the audit trail, and route data-rights or retention work into human-reviewed workflows.
Support access should be enabled only when the clinic asks VoiceOps to investigate an issue. Every support change and data-rights action is audited.
Support access is closed. Open a timed window only when the clinic owner/admin approves service-provider help.
No clinic audit events yet.
Track patient export, correction, deletion, and retention work with human completion, external-system boundaries, and auditable patient-data handling.
Retention review compares patient records with the clinic retention window. It flags candidates for staff review; it does not auto-delete medical or operational data.
Track queued messages, provider callback proof, dry-run evidence, consent-safe manual handoff, and failed-message recovery from one operating view.
Production delivery status comes from WhatsApp/provider webhooks. Demo delivery actions and manual status overrides are disabled.
Connect Mode foundation now shows connector status, source-of-truth controls, external mappings, and sync error review. Migration Mode is the next phase.