01Flagship innovation
Unified Health Intelligence Network
India has built the identity and consent layer for digital health at national scale. UHIN is the connective layer on top of it: an ABDM-compliant network that turns records held in many places into continuous care for one person.
02The national foundation
ABDM and ABHA, in short
The Ayushman Bharat Digital Mission is the Government of India's programme, run through the National Health Authority, to make health data portable. It is not one application: it is a set of building blocks any compliant system can plug into. These four are the ones UHIN is built on.
-
ABHA: one health identity
A 14-digit Ayushman Bharat Health Account, created from an Aadhaar number, a driving licence or a mobile number, that lets records made in different facilities be recognised as one person's. Holding one is voluntary.
-
Registries: verified facilities and professionals
The Health Facility Registry and the Healthcare Professionals Registry give every record a verified place and an accountable author, instead of a free-text hospital name.
-
Consent: brokered, never private
The Health Information Exchange and Consent Manager sits between the holder of a record and whoever requests it. Access is purpose-bound, time-bound, revocable and logged.
-
Standards: a shared clinical language
FHIR R4 for exchange, with SNOMED CT, LOINC and ICD-10 for meaning — so a result read in one system means the same thing in another.
03How we comply
Built against the specification, not adapted to it
Compliance is an architectural decision taken at the start, not an integration project bolted on at the end. This is the path UHIN is engineered against, and the commitments that follow from it.
- 01
Register the HIP and HIU roles
UHIN is built to operate as both a Health Information Provider and a Health Information User, integrated through the ABDM sandbox and promoted to production only after the integration milestones for each role are completed.
- 02
ABHA-first identity
Creation, verification and record linking are built around the ABHA number and address, including assisted enrolment paths for patients who will not complete it on a phone.
- 03
No movement without a consent artefact
The consent manager is a hard dependency, not a setting: if there is no valid artefact for the stated purpose and period, the request fails closed rather than falling back.
- 04
FHIR R4 as the internal model
Records are stored in the exchange model rather than translated into it on the way out, with SNOMED CT, LOINC and ICD-10 bound at ingest so meaning survives the journey.
- 05
Every record attributable
Entries resolve to a facility in the HFR and an author in the HPR, so provenance is a property of the record instead of a matter of trust.
- 06
Security and data protection by construction
Encryption in transit and at rest, tokenised identifiers, least-privilege access per role, a complete audit trail, and Digital Personal Data Protection Act 2023 obligations enforced at the data layer.
- 07
Alignment with the Health Data Management Policy
Retention, minimisation, grievance redressal and data residency in India are treated as design constraints on the architecture, not as policy documents beside it.
- 08
Conformance kept in the build
ABDM conformance cases run in continuous integration alongside the test suite, so a regression against the specification fails a build rather than a certification review.
This is the compliance path UHIN is engineered against. Altrix Labs does not claim completed ABDM certification, and nothing here implies endorsement by or affiliation with the National Health Authority.
04How it is put together
Four layers, one governed path
Producers write in, consumers read out, and every path between them is governed by the ABDM spine and the consent engine. Each layer can be replaced without disturbing the ones above or below it.
- 01
ABDM spine
HIP and HIU roles, ABHA resolution and FHIR R4 exchange, implemented against the national specification rather than retrofitted to it later.
- 02
Interoperability bus
Event streaming, an API gateway and the consent engine — the layer that decides whether a request is allowed before it ever reaches data.
- 03
Secure health data vault
A FHIR-native clinical store, encrypted per patient and federated by purpose, so a claims query cannot reach clinical detail it has no basis to see.
- 04
Intelligence layer
Decision support, population analytics and the data bank, all reading through the same consent and governance rules as any other consumer.
A federated vault, separated by purpose
Stores are isolated by purpose and access is granted per role, so a claims query cannot reach clinical detail it has no basis to see.
Clinical
Encounters, prescriptions, laboratory results and imaging.
Claims
Coverage, pre-authorisation and adjudication.
Pharmacy
Dispensing, medication history and demand.
Analytics
De-identified only, released under governed approval.
Identity
Personal identifiers, tokenised at the point of ingest.
05Who is in the network
Eight stakeholders, one record
A health network is not a product with users; it is a set of parties who each hold part of the truth about a patient. These are the parties UHIN connects, and what each of them brings to it.
-
Patients and citizens
The person the record belongs to, who grants and withdraws access to it.
-
Hospitals and clinics
Where most clinical documentation is created, and where a missing history costs the most time.
-
Clinicians
The people making the decision, who need the history at the moment of the consultation.
-
Diagnostics and laboratories
Results that become reusable evidence once they are coded and addressable.
-
Pharmacies
Dispensing and medication history, which is where interactions and adherence become visible.
-
Insurers and payers
Coverage, pre-authorisation and adjudication, currently decided with limited clinical evidence.
-
Government and public health
Programme delivery and population health, which depend on seeing need early enough to act.
-
Researchers and digital health platforms
The people who generate evidence and build services on top of the network.
06What each one gains
The advantage, stakeholder by stakeholder
Connection is only worth building if every party gets something they cannot get alone. Stated plainly, this is what each of them gets.
- Patients and citizens
- One portable history instead of a folder of paper; fewer repeated tests; the ability to see and revoke who holds access.
- Hospitals and clinics
- Admission with history attached, less time reconstructing a record, and referrals that arrive with evidence rather than a description.
- Clinicians
- Decision support that rests on a complete record, and a medication list that reflects every prescriber rather than one.
- Diagnostics and laboratories
- Results delivered in shared terminology and reused rather than reordered, plus a clear view of where capacity is genuinely missing.
- Pharmacies
- Reconciliation across prescribers, interaction checking with real history, and demand forecasting instead of guesswork.
- Insurers and payers
- Adjudication against clinical evidence, faster pre-authorisation, and fraud models built on observed utilisation.
- Government and public health
- District-level visibility of burden, coverage and gaps while budgets can still be moved, from consented data rather than manual returns.
- Researchers and digital health platforms
- Ethics-gated cohort discovery on de-identified data, and one governed integration surface instead of a separate deal per provider.
07The evidence
What the public record already shows
Every figure below is quoted from a named government source with the date it was stated. Together they describe both the scale ABDM has reached and the gap that remains for the stakeholders above.
- 0 crore
ABHA accounts created under ABDM
Press Information Bureau, Ministry of Health and Family Welfare · May 2026
- 0 crore
health records linked to an ABHA
Press Information Bureau, Ministry of Health and Family Welfare · May 2026
- 0
health facilities registered on the Health Facility Registry
Ministry of Health and Family Welfare, ABDM implementation update · 6 February 2025
- 0
healthcare professionals registered on the Healthcare Professionals Registry
Ministry of Health and Family Welfare, ABDM implementation update · 6 February 2025
- 0
hospitals empanelled under AB PM-JAY, of which 15,532 are private
National Health Authority, via Press Information Bureau · 1 December 2025
- 0 crore
hospital admissions authorised under AB PM-JAY, worth ₹1.60 lakh crore
National Health Authority, via Press Information Bureau · 1 December 2025
- 0%
of total health expenditure still paid out of pocket
National Health Accounts Estimates for India 2021–22 · released 2024
- 0%
government share of total health expenditure, up from 28.6% in 2013–14
National Health Accounts Estimates for India 2022–23 · released 2026
- 65%+
of India's population lives in rural areas, where the disease burden is highest
Ministry of Health and Family Welfare · as stated
Figures are quoted from the sources named beside them and were current at the dates shown; ABDM registry counts move continuously, and the live position is published on the National Health Authority's public ABDM dashboard.
08The case for the ecosystem
Why this has to be built
The identity layer is proven and the volume is real. What is missing is the connective tissue — and the cost of its absence is paid in repeated tests, lost continuity and decisions taken without the evidence.
-
The identity exists; the connection does not
Over nine hundred million health accounts and more than a billion linked records prove the identity layer works at national scale. What is still thin is the layer that turns those records into continuous care for one person moving between providers.
-
Rural continuity is a routing problem
Two-thirds of the population lives in rural areas, while specialist care is concentrated elsewhere. A record that follows the patient turns a referral into a handover, and a teleconsultation into a clinical conversation rather than a description of symptoms.
-
Urban care is fragmented, not scarce
In cities the problem inverts: many providers, each holding a fragment, none obliged to speak to the others. The cost shows up as repeated diagnostics and decisions made without the history.
-
Out-of-pocket spending is where the waste lands
Around two-fifths of health spending still comes directly out of households. Duplicate investigations are a large and avoidable part of that, and they are avoidable only if the earlier result can be found.
-
Planning still runs on assumption
Programme decisions are made on manual returns and last year's numbers. A governed, de-identified evidence layer is what lets a district be resourced on what is actually happening in it.
-
Intelligence is only as good as the record
Clinical AI trained and run on partial records inherits their gaps. Connection is the precondition for the intelligence, not a feature that comes after it.
09The health data bank
Evidence every stakeholder can decide on
A connected network produces something no single institution can hold alone: a governed, de-identified picture of how care is actually delivered.
-
Hospitals and health systems
Capacity and service-line planning grounded in real referral and utilisation patterns rather than last year's assumptions.
-
Government and public health
Disease burden, programme coverage and service gaps visible at district level, early enough to direct resources rather than report on them.
-
Insurers and payers
Risk and fraud models built on observed utilisation, so pricing and adjudication rest on evidence instead of proxies.
-
Pharmacies and diagnostics
Demand and stock-out forecasting, and a clear view of where diagnostic capacity is missing rather than merely under-used.
-
Research institutions
Cohort discovery and feasibility under ethics approval, without exposing an identifiable record to do it.
-
Clinicians and providers
Outcome and practice benchmarking against comparable peers, which is only meaningful once the data is comparable.
What the data bank is not
The guardrails are the reason it can exist at all.
- Identifiers are tokenised at ingest; the bank never holds a directly identifying record
- Release is aggregate and de-identified, with small cohorts suppressed rather than published
- Every release is purpose-bound and approved; research access is ethics-gated
- No individual record is sold, and no output is used for commercial targeting of a patient
- Consent and Digital Personal Data Protection Act 2023 obligations are enforced at the data layer, not by policy alone
10Collaboration
This is not something to build alone
A national health network is only as real as the institutions that join it. We are actively seeking to collaborate with state and central government, and with the digital health providers already holding part of the record.
-
State governments
Health departments and state digital health missions, where programme delivery and district-level need actually meet.
-
Central government
National programmes and the National Health Authority, aligning to ABDM as it is built rather than retrofitting to it later.
-
Digital health providers
Platforms, HIPs, HIUs, EMR vendors and health-tech builders already carrying part of the patient's record.
-
Provider networks
Hospital groups, diagnostic chains and pharmacy networks willing to connect a first corridor and prove it works.
11Get in touch
Connect a first corridor with us
Whether you run a state programme, a hospital network or a platform already holding part of the record — the useful conversation starts with one real corridor, not a national rollout.