At a glance

AttributeDetail
What it isAn enterprise master patient index and identity resolution layer for healthcare
What it doesCollapses duplicate and fragmented patient records into one governed identity
Sits on top ofYour existing EMR, HIE, data warehouse, and MDM investment
Reads fromEMR, labs, pharmacy, claims, ADT feeds, FHIR APIs, SDoH, consumer and device data
Matching methodsDeterministic, probabilistic, and referential, combined and weighted
TimingReal-time match check at registration, not an overnight batch job
OutputOne enterprise ID, a live crosswalk to every source ID, and a Patient 360 timeline
GovernanceReversible merges, threshold-based human review, full audit trail per decision
DownstreamBI dashboards, conversational assistants, and autonomous care agents
ComplianceHIPAA compliant, PHI stays inside your infrastructure boundary

The problem

One patient. Multiple systems. Zero reconciliation.

Registration typos create duplicates. Lookalike names cause overlays. When labs, pharmacies, and EMRs don’t know they’re describing the same patient, the gaps remain. The result: misidentification, denied claims, repeated tests, and registration lines that shouldn’t exist.

What is MPI and how it works?

An MPI gives each patient one verified identity inside a single system. An EMPI does the same across every system in the network, so the lab, the pharmacy, and the EMR are finally reading from the same patient record.

The Inferenz MPI pipeline runs end to end. Nothing skips ahead.

#1
Integrate

Raw feeds land from every source that touches a patient: labs, pharmacy, claims, EMRs, social determinants, consumer devices, third-party data. Clinical and non-clinical alike. This outlines patient data integration as part of the intake process.

#2
Standardize

Every record gets parsed, coded, and normalized into one format, using FHIR where it exists and HL7, ADT where it has to. A name splits into prefix, first, middle, last, and suffix instead of sitting in one text field. A date of birth gets forced into one format instead of three. An address breaks into street, unit, city, state, and ZIP+4. This fixes the shape of the data, not who it belongs to.

#3
Reconcile

Identities get matched, merged, and governed into one record. "Bob" and "Robert" become one name. A maiden name links to a married one. An enterprise ID is assigned and a crosswalk is built back to every local ID each source system uses. One patient becomes one record here, or nothing after it can be trusted.

#4
Enrich

Quality checks, gaps-in-care flags, and risk signals layer onto the unified record. None of that scoring means anything on a fractured record, which is why Enrich never runs before Reconcile.

#5
Activate

The resolved, enriched data feeds dashboards, conversational assistants, and AI agents. Every downstream answer traces back to Phase 3.

Inside the matching engine

Three methods, weighted, with a human in the middle band

1. Deterministic matching
Exact agreement on defined fields like name and date of birth

 

2. Probabilistic matching
Weighted likelihood across partial, imperfect field agreement

 

3. Referential matching
Comparison against a continuously updated third-party identity database

From one verified record to Patient 360

Reconcile answers "who is this patient?" Patient 360 answers "what happened to them, and when?"

Every system generates its own event with its own timestamp, all describing one person; a Patient 360 view converts each into one standard structure: what happened, when, and how it links to the rest. 

Patient identity resolution establishes who the patient is. Healthcare master data management resolves everything else in the frame, provider, location, billing code, so the Patient 360 timeline stays attached to the right patient, the right encounter, the right care team. 

What Patient 360 delivers, by care setting

Home health

  • Patient overview: demographics, active orders, and care team in one screen
  • Clinical and risk signals: vitals, medication changes, and risk score, tracked together
  • Patient journey: every visit and escalation, in order, from referral to discharge

Home care

  • Client overview: service plan, authorized hours, and care level in one view
  • Caregiver activity: visit history and tasks tied to the client record
  • Service utilization: hours delivered versus authorized, flagged before a plan lapses

Hospice

  • Patient and family overview: care plan and goals of care, visible to the whole rotation
  • Symptom and comfort trend: pain scores and interventions across visits
  • Visit and compliance history: required touchpoints against what actually happened

Hospitals

  • Encounter history: every admission, transfer, and discharge, linked to one chart
  • Diagnostic and lab trend: results from every department, read as one timeline
  • Care team coordination: who touched the case, when, and what changed

Business impact

One identity, fixed once. Every outcome, fixed downstream.

Improved revenue

  • Fewer denials at the source.
  • Faster clean-claim submission.
  • Measurable recovery of revenue lost to misidentification.

Reduced clinical risk

  • One complete clinical picture at the point of care.
  • Fewer decisions made on incomplete charts.
  • Lower adverse-event exposure.

De-risked enterprise growth

  • Cleaner M&A due diligence.
  • Stable EHR migrations, one identity every acquired facility reconciles against instead of a fresh mess each time.

Improved patient satisfaction

  • Real-time matching at registration.
  • Smoother throughput across departments, less time lost to manual chart lookups.

Robust governance

  • Governed merge decisions with a named owner.
  • A threshold, and a full audit trail defensible under HIPAA.

Trusted AI and analytics

  • An accurate denominator for population health and value-based care reporting.
  • Conversational AI that answers with one correct record
  • Agents that can act based on real charts.

Security and compliance

HIPAA compliant, deployed inside your environment

PHI never leaves your infrastructure or your cloud tenancy as part of patient identity management

Role-based access aligned to your existing clinical and data permissions

Every match, merge, split, and override logged with the score and rule that produced it

Complete audit trail exportable for HIPAA review and malpractice defence

The engine proposes. A named data steward decides. Authority does not move.

One patient. One record. One shot at getting it right.

Bring us one messy source pair. We'll show you the match rate, the overlay risk, and the merged records sitting inside it, before you commit to anything.

Book a 20-minute demo on your data