SCRIMED Healthcare Intelligence OS
A secure operating system, trust layer, workflow engine, and intelligence fabric for healthcare AI.
SCRIMED is advancing from individual healthcare AI features into a governed intelligence infrastructure layer with Agent Runtime, Clinical Knowledge Graph, Validation and Trust Lab, protected workspaces, model routing, and sovereign deployment readiness. Current execution remains synthetic and enterprise-evaluation only.
Current boundary
SCRIMED is productized for governed synthetic pilots, not live autonomous care.
SCRIMED is currently a governed synthetic pilot and enterprise evaluation operating layer. It does not ingest live PHI, autonomously diagnose, autonomously treat, submit claims, route patients, or execute production clinical workflows.
01framework: Next.js App Router with typed server components and route handlers
02database: Supabase Auth, Supabase Postgres, and Postgres row-level security for protected pilot workspaces
03auth: Supabase Auth bearer-token verification for tenant-admin protected pilot operations
04rateLimit: Upstash Redis rate limiting with bounded in-process fallback for protected mutations
05deployment: Vercel-hosted SCRIMED product app connected from the official Wix website
06dataPosture: Synthetic-only public product, demo, pilot, and evidence surfaces; live PHI remains blocked
Strategic phase plan
Build sequence from core operating layer to global healthcare intelligence fabric.
foundation-contract
Agent Runtime, Clinical Knowledge Graph, Validation and Trust Lab
Make SCRIMED's orchestration, evidence, knowledge, and validation contracts first-class before any live execution.
SCRIMED Agent Runtime, Clinical Knowledge Graph, Validation and Trust Lab, Health Records Safety Exchange
- Licensed clinician validation before clinical scoring is used in care settings.
- Regulatory intended-use review before clinical decision-support claims.
- Tenant identity, consent, BAA/legal path, durable audit, and approved connector controls before PHI.
foundation-contract
Persistent Workspace, Multi-Model Router, Sovereign Deployment
Enable resumable enterprise workspaces, vendor-neutral AI routing, and deployment profiles for controlled customer environments.
Persistent Agent Workspace, Multi-Model Router, Sovereign Deployment Mode
- Durable workspace state, idempotency, replay, retries, and failure quarantine for production tasks.
- Provider BAAs, regional processing rules, PHI sensitivity policy, fallback testing, and rollback controls.
- Private cloud, hospital-controlled, government, or edge deployment validation where customer data cannot leave the environment.
planned
Clinical Intelligence OS, Risk Horizon Engine, Population Intelligence
Build care-journey, preventive-risk, and population intelligence surfaces after trust, evidence, and governance are verified.
Clinical Intelligence OS, Risk Horizon Engine, Population Intelligence Layer
- Clinical validation studies, buyer baseline measurement, data-quality analysis, and clinician governance approval.
- No risk prediction, diagnosis, treatment, or patient instruction without authorized human review.
planned
Project Ark Healthcare Intelligence Fabric
Unify interoperable healthcare intelligence across organizations, regions, workflows, models, and governance systems.
Project Ark Healthcare Intelligence Fabric
- Enterprise architecture review, sovereign deployment approvals, formal security program, external clinical/regulatory review, and customer deployment evidence.
Agent Runtime
Shared memory, permissions, audit, planning, tool selection, error boundaries, cost, and latency become platform primitives.
AgentOS remains governed and synthetic-first while the runtime hardens toward long-running tenant workspaces.
Shared memory fabric
AgentOS defines session, operational, and knowledge memory with retention, RBAC, and prohibited-data boundaries.
- Durable tenant memory, deletion, residency, consent, and PHI minimization approval.
Shared permissions
AgentOS defines role permissions for admins, clinicians, RCM reviewers, and runtime services.
- Customer IdP, tenant roles, patient-context authorization, and break-glass policy.
Shared audit logs
AgentOS, TrustOS, workflow denial, and protected workspace layers expose auditable metadata events.
- Immutable durable storage, access review, retention, legal hold, alerting, and incident response.
Task decomposition and tool selection
Planner, Router, Specialist, TrustQA, and Governance agents produce synthetic task plans and denied production requests.
- Approved tool registry, credential scopes, retries, timeout policy, sandbox isolation, and live connector gates.
Cost and latency tracking
TrustOS model-route and observability contracts expose cost, latency, trust, override, and escalation signals.
- Provider telemetry, budget limits, tenant reporting, alerting, and fallback runbooks.
Clinical Knowledge Graph
FHIR, HL7, DICOM, SNOMED CT, ICD, LOINC, RxNorm, CPT, HCPCS, payer, pharmacy, and exchange standards form the reasoning substrate.
Current graph work is a typed foundation and synthetic contract. Live clinical graph ingestion requires customer authorization, PHI controls, terminology governance, and clinical validation.
standard
FHIR R4 and US Core
Resource model for patient context, observations, conditions, medications, procedures, encounters, claims-adjacent context, and care plans.
Interoperability control plane and synthetic conformance evaluations.
- Customer-specific FHIR profile validation, SMART authorization, consent, purpose-of-use, audit, and partner acceptance.
standard
USCDI and TEFCA
U.S. core data-class mapping and nationwide exchange governance context for record extraction, buyer data coverage, and future exchange readiness.
Health Records Safety Exchange and Interoperability registry readiness.
- Customer-specific data-class mapping, participant/exchange authority, privacy/security review, purpose-of-use, consent, and audit approval.
standard
HL7 v2
Event and message patterns for ADT, orders, results, scheduling, and legacy hospital interoperability.
Standards registry and future connector contract boundary.
- Interface-engine mapping, message validation, monitoring, reconciliation, and hospital integration testing.
standard
DICOM and DICOMweb
Imaging study, series, metadata, and retrieval patterns for radiology and imaging intelligence workflows.
Synthetic DICOMweb conformance planning and interoperability registry.
- PACS/VNA authorization, de-identification policy, imaging viewer controls, and radiology governance review.
standard
SNOMED CT
Clinical concept terminology for diagnoses, findings, procedures, body structures, and care concepts.
Knowledge-graph binding contract.
- Terminology license, version governance, mapping quality, and clinician terminology review.
standard
ICD-10 and ICD-11
Diagnosis and classification systems for clinical, reporting, quality, and global deployment context.
Knowledge-graph binding contract for claims-aware and population workflows.
- Coding compliance review and jurisdiction-specific coding/version governance.
standard
LOINC
Laboratory and clinical observation codes for longitudinal labs, vitals, and measurement context.
Knowledge-graph binding contract for risk, care-gap, and evidence workflows.
- Lab feed mapping, unit normalization, abnormal-flag validation, and data-quality controls.
standard
RxNorm
Medication normalization for ingredients, clinical drugs, dose forms, and medication-safety context.
Knowledge-graph binding contract for medication and risk-horizon workflows.
- Medication reconciliation policy, formulary integration, interaction-source governance, and pharmacist/clinician review where applicable.
standard
CPT, HCPCS, X12, NCPDP, and IHE profiles
Procedure, billing, payer, pharmacy, claims, and cross-enterprise document exchange context.
Interoperability and reimbursement-awareness planning.
- Payer policy review, coding/billing expert review, transaction testing, and no-guarantee reimbursement controls.
standard
CMS prior authorization APIs
Payer, provider, and prior-authorization API readiness for coverage context, missing documentation, and human-reviewed evidence packets.
Health Records Safety Exchange payer/prior-authorization extraction planning.
- Payer/trading-partner approval, API testing, coding review, legal/customer release authority, and no autonomous payer submission.
Graph contracts
SCRIMED graph contracts bind clinical, operational, payer, research, imaging, and outcome context to evidence and review state.
node
patient-context
Represent approved synthetic or tenant-authorized patient context without storing unrestricted clinical free text by default.
FHIR Patient, FHIR Encounter, FHIR Observation, FHIR Conditionnode
diagnosis-or-condition
Bind reviewable clinical concepts to terminology and evidence without creating autonomous diagnosis authority.
FHIR Condition, SNOMED CT, ICD-10, ICD-11node
medication
Normalize medication context for reviewable safety, adherence, and care-plan workflows.
FHIR MedicationRequest, FHIR MedicationStatement, RxNormnode
lab-or-vital
Track longitudinal observation context, units, reference ranges, and missing-data signals.
FHIR Observation, LOINC, ISO/IEEE 11073node
imaging
Represent imaging studies and metadata for imaging workflow context without autonomous imaging interpretation.
DICOM, DICOMweb, FHIR ImagingStudynode
procedure-or-service
Connect procedures, authorization needs, documentation requirements, and claims-aware context.
FHIR Procedure, CPT, HCPCS, SNOMED CTnode
guideline-policy-or-evidence
Attach evidence provenance, guideline versions, policy documents, validation timestamps, and source attribution.
Atlas Evidence Layer, Trust Card, Policy source registrynode
trial-or-research-protocol
Support reviewable TrialCore matching, eligibility evidence, missing criteria, and research operations workflows.
FHIR ResearchStudy, FHIR ResearchSubject, ClinicalTrials.gov-style metadatanode
payer-claim-or-authorization
Support revenue, prior authorization, denial risk, and reimbursement-awareness workflows under human review.
X12, FHIR Claim, FHIR Coverage, Da Vinci-aligned patterns where applicablenode
care-plan-or-outcome
Represent approved care-plan context, follow-up tasks, quality measures, and outcome signals after governance review.
FHIR CarePlan, FHIR Goal, FHIR Task, FHIR MeasureReportClinical Data Fabric
Healthcare source contracts, semantic normalization, provenance, and health-graph projection stay governed before any agent receives context.
SCRIMED Clinical Data Fabric is a no-live-PHI control plane for source contracts, semantic normalization, provenance, governance, and health-graph projection rules. It does not ingest live records, store PHI, expose raw schemas to agents, activate production connectors, submit payer transactions, mutate EHRs, interpret imaging, prescribe, diagnose, treat, contact patients, or approve customer go-live.
Source contracts
9 source contracts cover FHIR, HL7, DICOM, X12, documents, pharmacy, device, genomics, pathology, scheduling, portal, and patient-access context.
- Data boundary: no-live-phi-control-plane
- Connector authority: not-production-connector-approved
Semantic layer
6 semantic mappings normalize patient, condition, lab, medication, imaging, and claim concepts with provenance and confidence requirements.
- Agent data authority: semantic-layer-only-no-raw-schema-access
- Live ingestion authority: blocked-pending-customer-authorization
Health graph controls
22 node contracts and 10 edge contracts preserve source lineage, reviewer state, blocked uses, and human review requirements.
- Clinical care authority: not-authorized-live-care
- Blocked claims: 12
Clinical Data Governance
Every healthcare context request must pass deterministic policy checks before agents, tools, models, or connectors receive authority.
SCRIMED Clinical Data Governance evaluates metadata-only policy requests for purpose of use, role, data class, consent, tenant scope, minimum necessary access, destination, human review, residency, and contract readiness. It does not ingest live records, store PHI, approve production connectors, mutate records, submit payer transactions, contact patients, authorize external model PHI processing, diagnose, treat, prescribe, or approve customer go-live.
Policy engine
12 rules evaluate role, purpose, data class, action, destination, consent, tenant scope, minimum necessary access, review, residency, and contract readiness.
- Policy version: scrimed-clinical-data-governance-v2026-07-03
- Data boundary: metadata-and-policy-only-no-live-phi
Decision coverage
5 baseline control evaluations prove metadata access, deidentified review packets, external model PHI, record mutation, and semantic context decisions stay governed.
- Supported actions: 13
- Supported destinations: 9
Retained limits
Live care, production connectors, record mutation, patient outreach, payer submission, and external model PHI remain blocked.
- Clinical care: not-authorized-live-care
- Connector: not-production-connector-approved
- Record mutation: not-authorized
Clinical Context Gateway
Agents receive governed semantic envelopes, not raw schemas, raw connector payloads, or patient records.
SCRIMED Clinical Context Gateway converts authorized metadata-only requests into governed semantic context envelopes for agents. It does not accept raw patient records, expose raw database schemas, expose raw connector payloads, store PHI, activate production connectors, mutate records, submit payer transactions, contact patients, diagnose, treat, prescribe, interpret imaging, or approve customer go-live.
Semantic envelope control
The gateway covers 9 context scopes and9 source contracts with policy-cleared, metadata-only context envelopes.
- Envelope version: scrimed-context-envelope-v1
- Data boundary: governed-semantic-context-only-no-live-phi
Gateway proof coverage
5 baseline evaluations prove allowed, review-gated, unsafe-destination, live-data, and unregistered-source requests remain deterministic.
- Gateway controls: 13
- Raw schema access: blocked
Agent containment
Context delivery preserves source-contract provenance, confidence inputs, evidence requirements, blocked-use instructions, and audit hashes without widening agent authority.
- Raw connector payload: blocked
- Record mutation: not-authorized
- Patient outreach: not-authorized
Validation and Trust Lab
Every recommendation-like output must carry scores, evidence, provenance, model route, reviewer state, and audit metadata.
Validation fields attach evidence, uncertainty, review state, model route, and audit metadata. They do not authorize autonomous clinical decisions.
clinicalCorrectnessScore
Measure whether an AI output aligns with approved clinical facts, guidelines, and workflow intent.
- Required field
- TODO: define and validate scoring rubric with licensed clinicians before use in clinical environments.
completenessScore
Check whether required context, missing evidence, review owner, and blocked actions are present.
- Required field
- Current use is synthetic fixture and proof-packet evaluation only.
safetyScore
Evaluate escalation, prohibited action, uncertainty, privacy, and harm-prevention controls.
- Required field
- TODO: align with clinical safety, security, privacy, legal, and regulatory review before production use.
confidenceScore
Express model, evidence, and workflow confidence with uncertainty preserved.
- Required field
- Confidence never replaces authorized human review.
sourceAttribution
Identify evidence source, guideline, protocol, policy, or fixture source used by the output.
- Required field
- External source currency and clinical appropriateness require governance review.
evidenceTrail
Preserve the trace from input context to source evidence, Trust Card, reviewer state, and audit event.
- Required field
- Trace capture is metadata-first and must avoid PHI until approved storage exists.
guidelineReferences
Attach guideline or policy versions and validation timestamps where applicable.
- Required field
- Guideline selection, licensing, and update cadence need clinical governance approval.
modelProviderUsed
Record provider, model route, fallback state, latency, and cost basis for auditability.
- Required field
- PHI sensitivity, BAA, regional, and rollback controls are required before production routing.
reviewerStatus
Show whether the output is draft, pending review, approved for synthetic use, escalated, or denied.
- Required field
- Reviewer approval is not a substitute for licensed clinical responsibility.
auditLog
Retain event metadata for governance, replay, diligence, and incident review.
- Required field
- Durable production audit storage requires retention, access, encryption, and incident-response approval.
Model routing and deployment
SCRIMED should route across approved providers and run where customer governance requires.
foundation-contract
OpenAI
General reasoning, agent planning, structured extraction, and governed workflow support where policy allows.
- BAA or approved data boundary
- model logging policy
- PHI sensitivity routing
- fallback test
planned
Anthropic
Long-context review, safety-sensitive drafting, and policy-heavy evidence summarization after vendor approval.
- vendor review
- regional policy
- cost and latency benchmark
- output validation
planned
Google Gemini
Multimodal and long-context tasks where customer deployment, privacy, and accuracy controls support use.
- vendor review
- multimodal safety review
- data residency assessment
- fallback route
planned
Open-weight and local models
Private-cloud, edge, sovereign, and low-data-movement deployments where local inference is required.
- model evaluation
- security hardening
- update process
- clinical validation
planned
Future healthcare-specific models
Specialized clinical, imaging, payer, or research workflows after regulatory and customer governance review.
- intended-use review
- clinical validation
- source governance
- monitoring and rollback
foundation-contract
SCRIMED-managed cloud
Fastest synthetic pilot and enterprise evaluation path through Vercel-hosted product surfaces.
- environment isolation
- secrets management
- audit retention
- security monitoring
planned
Private cloud
Health-system or payer-controlled cloud deployment with customer network, identity, and data controls.
- customer IdP
- private networking
- regional storage
- observability export
planned
Hospital-controlled environment
Clinical workflows where live data must stay inside the customer's controlled environment.
- on-prem connector boundary
- local audit
- least-privilege service auth
- downtime procedures
planned
Government or sovereign cloud
Jurisdiction-specific residency, procurement, security, and sovereignty requirements.
- regional compliance mapping
- sovereign identity
- approved model route
- local incident response
planned
Edge or on-prem
Low-latency, low-connectivity, or no-data-egress environments.
- local inference
- patching process
- secure sync
- hardware and physical security
Clinical workflow automation
SCRIMED can reduce administrative drag through draft-only, source-attributed, human-reviewed workflow support.
Clinical workflow automation is limited to synthetic, metadata-only, draft, queueing, readiness, and human-reviewed support until customer clinical scope, privacy/security/legal review, PHI authority, connector approval, clinical governance, monitoring, and go-live approval exist.
synthetic-ready
Pre-visit chart prep and gap review
Organize known problems, medications, labs, preventive-care gaps, prior authorization context, and missing data into a review queue.
Draft-only visit agenda, missing-evidence checklist, and reviewer-ready preparation packet for synthetic or approved sandbox records.
- Buyer: Clinicians, care-team managers, and clinic operators
- Patient safety: missing-data disclaimer, source attribution required, stale-result flag, clinician-review required, no diagnosis or treatment recommendation
- Patient engagement analysis: access friction, missed appointment context, open follow-up need, portal-readiness signal
- Interoperability: FHIR Patient, FHIR Encounter, FHIR Observation, FHIR Condition, USCDI
- Clinician burden reduction: reduce manual chart hunting, compress agenda preparation, turn scattered evidence into a single review list
- Operations optimization: visit-readiness score, missing-record queue, care-team handoff queue
- Proof routes: /healthcare-intelligence-os, /health-records, /clinical-care-activation
- Blocked actions: live chart pull, patient-specific triage, clinical diagnosis, treatment plan recommendation
- Before live: Customer clinical scope, BAA/DPA where applicable, source-system connector approval, clinician rubric, and patient-context authorization.
- Boundary: Pre-visit automation is preparation support only; SCRIMED does not provide medical advice, diagnosis, treatment, or live chart execution.
design-contract
Documentation draft and clerical reduction
Convert encounter context into draft note structure, open questions, source references, and missing documentation prompts.
Draft note scaffold, evidence trail, and coding-adjacent clarification queue without final note signing, EHR filing, or clinical authorship transfer.
- Buyer: Physicians, APPs, scribes, documentation leaders, and operations executives
- Patient safety: draft-only watermark, clinician author remains responsible, source/evidence trail, hallucination and unsupported-claim check, no autonomous note finalization
- Patient engagement analysis: education-topic need, follow-up clarity need, language-access prompt, instruction-comprehension risk
- Interoperability: FHIR DocumentReference, C-CDA, LOINC document codes, SNOMED CT
- Clinician burden reduction: reduce after-hours documentation backlog, standardize note skeletons, surface incomplete documentation before signoff
- Operations optimization: documentation turnaround time, open clarification queue, note quality review queue
- Proof routes: /atlas, /health-records, /qa-claim-guard
- Blocked actions: autonomous documentation finalization, EHR filing, clinical authorship transfer, coding finalization
- Before live: Customer documentation policy, licensed clinical review, EHR workflow authorization, audit retention, and clinician acceptance criteria.
- Boundary: Documentation support remains draft-only and reviewer-gated until customer-approved clinical documentation controls exist.
customer-sandbox-required
After-visit follow-up readiness
Identify follow-up tasks, open referrals, lab/result review needs, education topics, and scheduling friction for human-owned workqueues.
Generate no-PHI workqueue plans, draft task categories, and escalation checklists; do not contact patients or route urgent care issues.
- Buyer: Care coordinators, population health teams, patient access, and ambulatory operations
- Patient safety: urgent/emergency boundary, human outreach approval, patient instruction blocker, task owner required, closed-loop follow-up audit
- Patient engagement analysis: follow-up completion risk, transportation/access friction, portal enrollment gap, communication preference readiness
- Interoperability: FHIR Task, FHIR CarePlan, FHIR CommunicationRequest, USCDI
- Clinician burden reduction: reduce manual follow-up queue sorting, convert care-plan fragments into worklists, separate administrative follow-up from clinical escalation
- Operations optimization: follow-up backlog, handoff completeness, open-loop referral count
- Proof routes: /clinical-care-activation, /health-records, /service-delivery
- Blocked actions: patient outreach, urgent triage, clinical instruction, care-plan mutation
- Before live: Customer outreach policy, consent/communication governance, emergency escalation protocol, care-team ownership, and connector approval.
- Boundary: SCRIMED may analyze engagement readiness and draft workqueues, but it does not contact patients or issue clinical instructions.
synthetic-ready
Referral, prior authorization, and documentation workbench
Map order/referral context, payer policy evidence, missing documentation, deadlines, and human reviewer requirements.
Prepare evidence packet outlines, policy checklists, and status workqueues without payer submission, coverage determination, or claim guarantee.
- Buyer: Referral teams, revenue-cycle leaders, prior authorization staff, and specialty access operators
- Patient safety: coverage-decision blocker, payer-submission blocker, policy source attribution, human reviewer approval, no reimbursement guarantee
- Patient engagement analysis: authorization delay risk, access bottleneck, missing patient document request, care access friction
- Interoperability: FHIR Coverage, FHIR Claim, FHIR ServiceRequest, X12 278, CMS prior authorization APIs
- Clinician burden reduction: reduce manual policy lookup, package missing documentation requests, separate clerical payer tasks from clinician review
- Operations optimization: authorization cycle time, missing documentation rate, referral leakage signal
- Proof routes: /health-records, /interoperability, /offerings
- Blocked actions: payer submission, coverage determination, appeal filing, reimbursement guarantee
- Before live: Payer/trading-partner approval, customer policy source authority, coding review, legal review, and human submission workflow.
- Boundary: Prior-authorization automation is evidence preparation only; SCRIMED does not submit claims or guarantee payment.
external-review-required
Medication reconciliation safety review
Surface duplicate-medication context, missing medication history, reconciliation gaps, allergy references, and source discrepancies for qualified review.
Create safety-review prompts and discrepancy queues without medication advice, interaction claims, prescribing, or patient instructions.
- Buyer: Clinical pharmacists, primary-care teams, care managers, and quality leaders
- Patient safety: pharmacist/clinician review required, medication-change blocker, source discrepancy flag, interaction-source governance required, patient instruction blocker
- Patient engagement analysis: adherence conversation need, medication access friction, pharmacy follow-up gap, education support need
- Interoperability: FHIR MedicationRequest, FHIR MedicationStatement, FHIR AllergyIntolerance, RxNorm
- Clinician burden reduction: reduce medication list comparison work, highlight missing reconciliation evidence, route discrepancy queues to the right reviewer
- Operations optimization: med-rec completion queue, high-risk discrepancy queue, pharmacy-review load
- Proof routes: /healthcare-intelligence-os, /clinical-authority-readiness, /qa-claim-guard
- Blocked actions: medication change recommendation, prescribing, drug interaction claim, patient instruction
- Before live: Medication-source licensing, pharmacist/clinician validation, customer medication policy, EHR reconciliation workflow, and monitoring.
- Boundary: Medication safety review is a qualified-review queue; SCRIMED does not prescribe, change medications, or instruct patients.
design-contract
Care-gap and population engagement analysis
Aggregate care gaps, quality-measure readiness, access friction, panel workload, and equity review signals for program planning.
Produce aggregate or synthetic engagement and workload dashboards, not patient-specific outreach lists or risk prediction.
- Buyer: Population health, quality, payer operations, and value-based care teams
- Patient safety: aggregation threshold, equity review, consent and purpose-of-use gate, no patient-specific scoring, anti-discrimination review
- Patient engagement analysis: care-gap closure friction, access equity signal, message-channel readiness, community resource need
- Interoperability: FHIR MeasureReport, FHIR Group, USCDI, TEFCA governance context
- Clinician burden reduction: reduce manual quality roster review, separate outreach planning from clinical decision work, prioritize operations bottlenecks before clinician escalation
- Operations optimization: care-gap backlog, panel workload, quality-measure readiness, equity-review queue
- Proof routes: /healthcare-intelligence-os, /operational-efficiency, /public-market-readiness
- Blocked actions: patient-specific scoring, automated outreach, risk prediction claim, quality outcome guarantee
- Before live: Data rights, aggregation policy, equity/bias review, customer governance, measure-source review, and outreach policy.
- Boundary: Population engagement analysis remains aggregate or synthetic until approved data rights and patient outreach governance exist.
customer-sandbox-required
Discharge transition workflow optimization
Map discharge tasks, pending results, medication reconciliation handoffs, referral status, post-acute needs, and readmission-risk review prompts.
Prepare transition checklist and handoff packet templates without discharge instructions, order entry, patient routing, or clinical risk scoring.
- Buyer: Hospitals, care-transition teams, case management, and post-acute coordinators
- Patient safety: pending-result flag, handoff owner required, no discharge instruction generation, readmission-risk claim blocker, post-acute referral authority gate
- Patient engagement analysis: transport or post-acute barrier, follow-up appointment friction, caregiver support need, home-instruction clarity risk
- Interoperability: FHIR Encounter, FHIR CarePlan, FHIR Task, HL7 v2 ADT, C-CDA
- Clinician burden reduction: reduce discharge checklist assembly, surface incomplete handoffs, route administrative barriers before clinician escalation
- Operations optimization: transition-readiness queue, pending-result handoff rate, post-acute bottleneck map
- Proof routes: /clinical-care-activation, /interoperability, /service-reliability
- Blocked actions: discharge instruction generation, order entry, post-acute referral submission, readmission prediction
- Before live: Hospital discharge policy, clinical governance, ADT/feed approval, post-acute workflow authority, and incident escalation model.
- Boundary: Transition workflow support is operational checklisting only; SCRIMED does not discharge patients or direct care.
synthetic-ready
Clinician inbox and administrative triage
Classify synthetic messages, refill/admin requests, paperwork tasks, scheduling blockers, and unclear clinical-intent signals into human-owned queues.
Draft routing labels, response templates, and clerical task groups without patient advice, refill approval, or clinical triage.
- Buyer: Clinicians, practice managers, access centers, and administrative operations leaders
- Patient safety: clinical intent escalation, urgent symptom blocker, patient advice blocker, human sender approval, message source audit
- Patient engagement analysis: message backlog, common confusion topic, access request pattern, paperwork friction
- Interoperability: FHIR Communication, FHIR Task, SMART App Launch, AuditEvent
- Clinician burden reduction: reduce inbox sorting load, separate clerical messages from clinical review, prepare human-send response drafts
- Operations optimization: inbox backlog, clerical deflection rate, clinical escalation queue, response-time variance
- Proof routes: /client-onboarding, /continuous-review-audit, /qa-claim-guard
- Blocked actions: patient medical advice, refill approval, urgent triage, message sending without human approval
- Before live: Customer messaging policy, identity and role gates, human-send workflow, emergency escalation boundary, and audit retention.
- Boundary: Inbox support stays administrative and draft-only; SCRIMED does not send messages or triage patients autonomously.
Future intelligence layers
Care-journey, risk-horizon, and population intelligence stay gated until validation and governance mature.
before-visit
Pre-visit chart review, missing-data detection, risk signal summary, and agenda drafting.
- Draft operational support only; no diagnosis, treatment recommendation, or patient instruction.
during-visit
Ambient listening support, topic tracking, note scaffolding, and patient education draft support.
- Clinician remains author and reviewer; no autonomous documentation finalization or advice.
after-visit
Draft documentation, coding support, care-plan drafting, follow-up queueing, payer and RCM support.
- No EHR filing, final coding, claim submission, patient outreach, or order entry without authorized review.
Heart failure
Surface missing-data, trend, medication, encounter, and follow-up review prompts in synthetic examples.
- Validated clinical model, cardiology governance, data-quality review, and escalation protocol.
Diabetes
Map A1c, medication, care-gap, monitoring, and access signals to reviewable workflow prompts.
- Endocrinology/primary-care review, evidence versioning, and patient-communication governance.
Chronic kidney disease
Map labs, medication safety, referral gaps, and follow-up context into human-reviewed risk workqueues.
- Nephrology validation, unit normalization, lab-feed governance, and care-team escalation.