Deployment roadmap
BrainMeBack sequences deployment to match clinical rigor with pilot validation. Phase 1 is the in-clinic, supervised, clinician-prescribed deployment model designed for initial pilot sites. Phase 2 is the home-extension layer — built, in the same codebase, hardening through pilot validation before broad release. There is no second product to build.
Phase 1 — pilot-ready
What initial pilot sites can deploy first. The full prescribing dashboard, clinic-iPad delivery, demonstration mode, and analytics capture pilots need to characterize outcomes — under the supervising clinician’s control at every step.
Multi-select plan assembly, severity-aware module gating, protocol presets for TBI / stroke / MCI / post-COVID, and a single reviewed-plan screen before anything is prescribed. Every change is logged with clinician identity and timestamp.
The clinician can run a module with the patient in the office on a clinic iPad — for a baseline at intake, supervised practice, or guided onboarding. Recorded sessions feed the same dashboard analytics as Phase 2 at-home use.
Walk the patient through how the app works without recording anything to their record. Designed so the patient sees the actual app surface before their first recorded session — onboarding friction goes to zero.
Real-time accuracy, reaction time, engagement, clinician-flagged session events, and per-exercise d-prime trends. Exportable in formats designed for pilot outcome characterization and survey review (TJC, CARF, internal audit).
Field-level AES PHI encryption, Postgres Row-Level Security for per-clinician + per-organization isolation, hash-chained audit log, 15-minute idle session timeout, SMART on FHIR OAuth 2.0 + PKCE for EHR connections where deployed. BAA-backed deployment is part of the enterprise readiness path.
Core cognitive, motor, ADL (Safe Steps), and arousal/focus/rest categories. Every module is severity-gated to the patient’s Rancho Los Amigos level — difficulty is capped, never exceeded.
Phase 2 — ready, hardening through pilot
The home-delivery layer is built into the same codebase as Phase 1. It hardens through pilot validation before broad release — the same clinician-supervised workflow and security posture as Phase 1, extended to the patient’s home. The protocol continues to be refined against the published cognitive-rehabilitation and attention-research literature.
The same prescription assembled in the clinician dashboard delivers to the patient’s take-home iPad. The arousal–focus–rest session rhythm runs identically; PHI stays encrypted at the field level; every session feeds the same clinician dashboard.
Caregivers invited by the clinician get a dedicated portal: their own login, role-scoped patient view, ability to support the patient’s home practice without seeing PHI they shouldn’t. Invitation, login, signup, and patient-view pages are all built; pilot validation hardens the consent and access patterns.
The ADL training program includes a built-in caregiver mode for at-home practice supervision — body-check pre-screening, environment-checklist walkthroughs, and per-step clinical imagery designed for a caregiver to coach the patient through 31 activities of daily living.
Integration points are scaffolded into the dashboard for clinician–patient session continuity between in-person visits. The framework is in place; vendor selection and production hardening proceed through pilot.
Between-visit home sessions feed the same dashboard as in-clinic sessions, building a longitudinal cognitive trajectory the supervising clinician can review across visits. The data flywheel for outcomes research is structural, not a future ambition.
Field-level PHI encryption, Row-Level Security, audit log, severity gating, CDS-aware regulatory framing — all of it applies identically when the prescription is delivered to a take-home iPad. Phase 2 is not a different product. It is the same product, in the home.
What stays continuous
The pilot phasing reflects clinical rigor, not engineering limits. Both phases operate under the same supervising clinician, the same regulatory framing, the same security architecture, and the same auditable prescription workflow.
Why the phasing
Phase 2 is not a future product waiting to be built. It is in the same codebase as Phase 1, operates under the same clinician supervision model, and inherits the same security architecture. The reason it sequences after Phase 1 is clinical, not engineering: home delivery to cognitively impaired patients deserves pilot-validated workflows, caregiver consent patterns hardened against real households, and outcome data characterizing what the at-home protocol actually delivers between visits.
Initial pilot sites can deploy Phase 1 with the full prescribing surface, supervised in-clinic delivery, and analytics they need to characterize outcomes. The same data characterizes the protocol so Phase 2 broad rollout rests on evidence rather than ambition.
For investors and forward-looking buyers: Phase 2 is the TAM-unlock story — the post-acute discharge → home → continued-care continuum that today is broken by the hand-off between clinic and household. The infrastructure to bridge that gap is built. Pilot validation is the unlock event, not the build event.
Scientific grounding
BrainMeBack’s clinical architecture and security posture rest on widely-documented frameworks — none of them ours. The list below names the published basis for the platform, independent of any individual researcher.
Published 1972 at Rancho Los Amigos National Rehabilitation Center, widely used worldwide in TBI rehabilitation. Drives severity gating across every module — exercises above a patient’s Rancho level are surfaced and excluded, never silently dropped.
Spaced-retrieval scheduling algorithm developed and published by Piotr Wozniak (SuperMemo World). Drives the consolidation schedule in memory-recall modules. Attributed in the site footer per the algorithm’s licensing terms.
International Dysphagia Diet Standardisation Initiative Framework, licensed under CC BY-SA 4.0. Drives texture-modification filtering in the Brain Fuel nutrition module so foods above a patient’s safe swallow level are auto-excluded.
Session structure — optional cyclic-sighing settling, focused practice, and a recovery interval — is mapped to the canonical attention networks (Posner & Petersen 1990; Raichle 2001). Focus-to-recovery intervals are clinician-prescribed within evidence-supported bounds.
EHR interoperability built on the SMART on FHIR specification with OAuth 2.0 + PKCE. Hospital deployments connect via the same standard Epic, Cerner, VA VistA, and Allscripts publish against. No proprietary integration layer between BrainMeBack and the EHR.
Security posture architected against the HIPAA Security Rule (45 CFR Part 164). Regulatory framing designed to support clinician-reviewed CDS workflows under Section 3060 of the 21st Century Cures Act (FD&C Act §520(o)(1)(E)).
Naming individual researchers and clinical collaborators is reserved for the moment their institutions have completed conflict-of-interest review and signed off on public attribution. Until then, the platform stands on the published frameworks above.
For pilot sites: a 30-minute walkthrough with your clinical lead. For investors, partners, and press: a 1:1 conversation directly with the founder.