100%
of the available chart read on every patient
250+
conditions detected across specialties
<10s
from chart open to complete assessment
Days
to go live at a new health system, not months
01 · Ingestion

Every chart. Every time. Completely.

Regard reads the full patient chart — labs, vitals, medications, notes, imaging reports, and more — every time a physician opens it. Structured data and unstructured clinical language are both understood natively, regardless of which EHR a hospital runs.

  1. 01

    Complete chart coverage

    Not a sample. Not a summary. Every available data point — labs, vitals, medications, procedures, notes, imaging reports — pulled fresh every time the chart is opened.

  2. 02

    Understands clinical language

    Clinical notes are parsed with full awareness of context — so "no history of CHF" and "admitted with CHF" are never confused. PDFs, HTML reports, and free-text all contribute to the clinical picture.

  3. 03

    Battle-tested integrations

    Live in health systems across the country, handling the edge cases and vendor-specific behaviours that only show up at scale. Years of production experience inform every integration.

  4. 04

    Days to go live, not months

    Onboarding a new health system is configuration, not custom development. The platform adapts to each hospital's environment without new engineering work.

Liveload 7f3a·c21Epic · FHIR R4triggered by chart open0.00s
  1. Queued
  2. Planned
  3. Fetching
  4. Merged
  5. Mapped
  6. Stored
0 resources · 0 documents parsed · 0 changed rows written · 14 loads in flight
 
02 · Clinical Understanding

One clinical language across every hospital.

Every hospital codes things differently — different systems, different shorthand, different local conventions. Regard resolves all of it into a single clinical understanding. When the system reasons about "heart failure," it doesn't miss cases because one hospital uses a different code than another, or because a physician wrote it differently in a note.

Clinical Concept Library · one slice of the graph57 of hundreds of thousands

Hover or tab through a concept to see the codes that resolve to it and how it connects.

ConditionLab or vitalMedicationDrug classFinding or procedureLine style encodes the relationship type
Platelet countThrombocytopeniaHeparin-inducedthrombocytopeniaImmunethrombocytopeniaHeparinWarfarinAnticoagulantsINRPlatelettransfusionHemoglobinAnemiaIron deficiencyanemiaFerritinBleedingLactateWhite cellcountProcalcitoninBlood cultureFeverSepsisSeptic shockPneumoniaMean arterialpressureAntibioticsPiperacillin-tazobactamCreatinineeGFRBUNUrine outputAcute kidneyinjuryChronic kidneydiseaseNephrotoxicagentsVancomycinIodinatedcontrastTroponinBNPEjection fractionAcute coronarysyndromeHeart failureAtrial fibrillationBeta blockersMetoprololLoop diureticsFurosemideGlucoseHemoglobinA1cAnion gapType 2 diabetesDiabeticketoacidosisHypoglycemiaMetforminInsulinSodiumPotassiumHyponatremiaHyperkalemiaSIADH
  1. 01

    Universal clinical vocabulary

    Every major clinical coding system — plus each hospital's proprietary codes and shorthand — resolved to a single canonical understanding. No diagnosis is missed because of a coding difference.

  2. 02

    Relationships, not just codes

    Regard understands that a drug belongs to a class, that a lab measures a specific thing, and that conditions have complications — so reasoning reflects how medicine actually works.

  3. 03

    Always current

    New codes, new medications, new tests — reflected immediately without downtime. The clinical knowledge base evolves continuously as medicine does.

  4. 04

    Hospital-specific without custom code

    Each site's local terminology is mapped once. After that, every piece of diagnostic logic works everywhere — no per-hospital engineering.

03 · Diagnostic Reasoning

Clinical reasoning you can trace and trust.

At its core, Regard reasons about patient data the way a careful physician would — considering the full clinical timeline, weighing evidence, and ruling out alternatives. Every condition identified comes with a complete evidence chain. No black boxes. No unexplainable scores. Every assessment can be audited by a physician, a CDI team, or a regulator.

Rule · Thrombocytopenia · adult inpatient · v14 · published
DECISIONThrombocytopenia → emit assessment
AGGREGATELowest platelet count in the last 24h → 38 × 10⁹/L
CHECKPlatelet count < 150 × 10⁹/L
SOURCEPlatelet count · CCL platelet_count
AGGREGATEBelow threshold on every segment in the window → true
DECISION TABLESeverity from the lowest value
100 – 149Mild
50 – 99Moderate
< 50Severe38 ✓
CHECKAcuity, first fell below threshold this admission → acute
CHECKNot better explained by a more specific rule → true
SOURCEHeparin on the active med list → false
SOURCEITP treatment on the active med list → false
Result · Thrombocytopenia, severe, acute. Four of four platelet results sat below 150 across 24 hours; the lowest was 38 × 10⁹/L at 04:12. Neither heparin nor ITP treatment is active, so no more specific rule takes precedence, neither heparin-induced nor immune thrombocytopenia. View the four source results →
Platelet count · last 24 hours
MildModerateSevere200100500threshold 150 × 10⁹/L142966138−24h−18h−12h−6hnowunknown

Each result stays valid until the next one supersedes it, so the rule evaluates a continuous series of segments rather than four isolated points. Before the first result there is nothing to evaluate, so the rule returns unknown across that interval, not false.

  1. 01

    Transparent by design

    Every conclusion traces back to specific evidence in the chart. Physicians can see exactly why a condition was identified — and disagree if the evidence doesn't hold.

  2. 02

    Clinically accurate over time

    Regard doesn't just look at a single lab value in isolation. It understands how findings evolve — distinguishing a transient spike from a sustained trend, and reasoning across the full stay.

  3. 03

    Handles missing data honestly

    Results distinguish a ruled-out condition from one the data can't judge, a distinction that most rule engines collapse and clinicians can't afford to lose.

  4. 04

    Consistent & reproducible

    The same chart always produces the same assessment. No variability between runs, no drift over time. Every result is auditable and can be explained to a regulator.

04 · Clinical Governance

Clinical expertise at the center.

Regard's diagnostic logic is written and maintained by clinicians — not engineers. Our clinical team continuously updates and expands condition coverage based on the latest guidelines, peer review, and real-world performance data.

  1. 01

    Clinician-authored

    Every condition in Regard is defined, reviewed, and approved by physicians. Engineering builds the platform; clinicians build the medicine.

  2. 02

    Peer-reviewed rigor

    Every piece of diagnostic logic goes through clinical peer review before reaching a single patient. Only approved versions are active — every prior version is archived for audit.

  3. 03

    AI-accelerated, clinician-governed

    AI helps our clinical team work faster — but clinicians decide what ships. No diagnostic logic reaches patients without physician approval.

  4. 04

    Continuously improving

    New conditions, updated guidelines, refined thresholds — all without software releases or downtime. The clinical knowledge base grows every week.

v14 Published · authored by Dr. J. Rivera · reviewed by 2 clinicians

Thrombocytopenia · Adult inpatient

SourcePlatelet count
Check< 150 × 10⁹/L
AggregateLowest value in the last 24 hours
Decision tableSeverity band · mild / moderate / severe
CheckNot better explained by heparin or ITP treatment
DecisionEmit severity + acuity
Ask AI: "Add a check for a fall of 25% or more over the last 48 hours."
Reliability, scale & compliance

Built to the audit standard healthcare demands.

Regard's technology stack is designed for the reliability, traceability, and compliance expectations of hospital customers and their regulators. Every fetch, every mapping, every diagnosis is accounted for.

  1. 01

    Audit-ready

    Every assessment, every data fetch, every version change — logged and traceable for compliance. Designed for the scrutiny of hospital regulators and payers.

  2. 02

    Always complete

    If a data source is temporarily unavailable, Regard retries automatically. No partial assessments reach clinicians — every evaluation runs against the full available chart.

  1. 03

    Built for scale

    Handling millions of chart loads across health systems without degradation. Infrastructure is designed to grow with clinical demand, not against it.

  2. 04

    Respectful of hospital systems

    Background work is scheduled to minimize impact on hospital infrastructure. Regard is designed to be a good citizen on the networks it connects to.

See how Regard reads a chart.

Bring a de-identified case. We will run it end to end and show you every step the assessment came from.