01
Patient command center
Single-pane patient profile that unifies medications, labs, PGx, wearables, phenotype, and timeline into a clinician-readable chart.
FHIR R4SMART-on-FHIR
PrecisionRx AI · Clinical Intelligence OS
One evidence-linked engine.
Many surfaces of practice.
For personalized medicine.
Genomic, lab, medication, wearable, phenotype, and EHR signals — unified into a clinician-readable chart and surfaced as the small set of decisions that actually matter for today's visit.
Modules9 included
StandardsFHIR R4 · LOINC · RxNorm
EvidenceCPIC / PharmGKB versioned
PostureSaMD-track / pre-clearance
Intended useSaMD-track / pre-clearance.
PrecisionRx AI is the explainable precision-medicine decision-support layer that plugs into the EHR a clinic already uses. It organizes patient-specific clinical data into a structured context, runs evidence-linked decision-support modules against that context — starting with pharmacogenomics and drug safety — and surfaces cited recommendations for a licensed clinician to review and approve.
PrecisionRx AI is decision support. It does not diagnose, prescribe, or act autonomously. The clinician makes every clinical decision.
Built forEight clinical specialties supported out of the box.
- 01PGx medication safety
- 02Cardiometabolic optimization
- 03Psychiatry medication review
- 04Renal / hepatic monitoring
- 05Medication tolerability monitoring
- 06Rare disease feasibility
- 07Polypharmacy review
- 08Preventive + longevity
Platform modules§ I
Nine modules.
One spine of evidence, provenance, and clinician review.
Every module routes back to the same versioned rule engine and audit-logged review surface. Adoption can start with a single lane and expand without re-platforming.
02
Pharmacogenomics engine
CPIC- and PharmGKB-aligned phenotype calls with versioned evidence. Drug-gene safety surfaces only when the data quality clears the bar.
CPIC Level APharmGKB 1A
03
Diagnostics intelligence
Lab + IVD normalization into FHIR-ready DiagnosticReport / Observation context with LOINC, source provenance, and quality grading.
LOINCHL7
04
Cardiometabolic optimization
ApoB / Lp(a) / HbA1c / lipid / CGM trends interpreted with statin tolerability, renal function, and pharmacogenomic context.
AHA / ACCADA
05
Medication tolerability monitoring
Sleep, HRV, resting heart rate, and CGM trends correlated to medication initiation events. Catches side-effect signals before the next visit. Wear-time confidence and stale-feed gating built in.
Apple HealthOuraGarmin
06
Advanced therapy explorer
Rare-disease feasibility, suspected mechanism, candidate categories for specialist review, and a missing-data checklist.
ClinVarOMIMHPO
07
Agentic workflow layer
Intake, evidence, visit-prep, report, referral, care-gap, and patient-explanation drafts. Model + prompt versioning. Drafts only.
Audit-logged
08
Clinician review + audit
Co-sign workflows, suppression, amendment, hash-chained audit log, quality-override governance, and rule-version snapshots.
IEC 62304
09
Report suite
Clinician PDF, EHR paste block, patient-friendly summary, and specialist referral templates. Provenance snapshot per report.
EHR-ready
Architecture§ II
Source data in.
Evidence-linked decisions out.
A composable pipeline that keeps the clinician at the keel. Every stage is auditable, versioned, and reversible.
- istage
Source systems
EHR, labs, PGx vendors, sequencing partners, wearables, patient questionnaires, PDFs.
- iistage
Intake + normalization
SMART-on-FHIR, HL7 v2 / FHIR R4, VCF, CSV, OCR. Source-tagged with provenance.
- iiistage
Patient context store
FHIR-shaped Patient / Medication / Lab / Variant / Phenotype / Timeline records.
- ivstage
Evidence knowledge base
Versioned CPIC / PharmGKB / FDA-label / guideline entries with aging review queue.
- vstage
Insight engine
Deterministic specialty-laned rules. Categories: safety, efficacy, monitoring, optimization.
- vistage
Agentic workflow layer
Intake / evidence / visit prep / report / referral / patient-explanation drafts.
- viistage
Clinician review + co-sign
Lifecycle states, required rationale, supervising-clinician co-sign on prescribing-relevant accepts.
- viiistage
Reporting + EHR export
Clinician PDF, EHR paste, patient summary, specialist referral. Source-data snapshot per report.
9Composable modules sharing one evidence engine and one audit log.
8Specialty review lanes — primary care, cardiometabolic, psychiatry, longevity, more.
0Autonomous prescriptions. Every recommendation requires clinician acknowledgement.
Specialty lanes§ III
The same engine, configured for the practice you actually run.
Switch the lane and the panel composition, default rules, and report templates reshape themselves. Same evidence backbone, different daily surface.
01
PGx medication safety
02
Cardiometabolic optimization
03
Psychiatry medication review
04
Renal / hepatic monitoring
05
Medication tolerability monitoring
06
Rare disease feasibility
07
Polypharmacy review
08
Preventive + longevity
Standards & vocabularies§ IV
Built on the standards your other systems already speak.
No proprietary data shapes for clinical concepts. Imports validate against industry vocabularies; exports return them.
- FHIR R4
- Patient · MedicationRequest · Observation · DiagnosticReport
- LOINC
- Lab and biomarker observations
- RxNorm
- Medication identification and interactions
- SNOMED CT
- Conditions, problem list, allergies
- HPO
- Phenotype findings and rare-disease referral
- HGVS / ClinVar
- Variant nomenclature and clinical significance
- CPIC / PharmGKB
- Drug-gene evidence linkage and phenotype calls
Bring it into your clinicSee the chart before
See the chart before
you book the demo.
Open a sample patient — Jane Smith, 38F, three high-priority items waiting — and walk through the workflow without scheduling anything.