On-premise deployment
Your hospital. Your servers.
Your data.
Nadi is installed on infrastructure you own and control. There is no shared cloud tenancy and no patient record ever leaves your premises. Every plan is the same on-premise product — they differ by the modules enabled and the scale they are built for.
Compute needs vary with patient volume, concurrent users and retention, so each deployment is sized and quoted with you rather than off a price list.
Core clinical and billing workflows, running in your clinic.
Best for
Clinics and nursing homes up to ~50 beds
Deployment
Single serverA modest footprint on one machine.
Scoped with our team · no self-serve signup
- Patient records & MRN
- OPD scheduling & consultations
- IPD admissions & wards
- Billing with GST 2.0
- Pharmacy & drug register
- ABDM M1 (ABHA creation & linking)
- NABH statutory registers
- Single-server Docker install
- Email support · 48 h SLA
Full clinical operations for a working hospital.
Best for
Hospitals of roughly 50–200 beds
Deployment
Single serverMore headroom for concurrent users across departments.
Scoped with our team · no self-serve signup
- Everything in Basic
- TPA & insurance claims
- OT management & scheduling
- Lab (LIS) & sample tracking
- Radiology (RIS) & PACS link
- Blood bank & CSSD
- Day care, HR & procurement
- ABDM M1 + M2 (e-Prescriptions)
- WhatsApp notifications · 8 h SLA
Multi-branch operations and full ABDM interoperability.
Best for
200+ bed hospitals and multi-branch groups
Deployment
Scales to multi-nodeScales out, with a high-availability database option.
Scoped with our team · no self-serve signup
- Everything in Pro
- Multi-branch management
- Custom clinical workflows
- ABDM M1 + M2 + M3 (HIE)
- FHIR R4 export & import
- PCPNDT Form-F automation
- 7-year audit log retention
- DPDP Act fiduciary contract · BAA
- Phone + WhatsApp · 2 h SLA
Everything, plus Asha — the clinical AI, running on your hardware.
Best for
Hospitals that want AI without sending data off-site
Deployment
Multi-node + GPUAdds GPU capacity so Asha can run entirely on your hardware.
Scoped with our team · no self-serve signup
- Everything in Premium
- ✦ Asha WhatsApp patient agent
- ✦ Asha TPA denial predictor
- ✦ Asha discharge summary AI
- ✦ Asha clinical decision support
- ✦ Asha demand forecasting
- Air-gapped AI option — no data leaves site
- Dedicated account manager · 1 h SLA
Estimate your plan
Tell us how your hospital runs and we'll point you at the right plan. Exact compute is worked out with you during scoping — this is about fit, not a spec sheet.
Recommended plan
- 80 beds fits a Pro deployment
- 40 named users exceeds a Basic install
- TPA and insurance claims start at Pro
Full feature comparison
All four plans install on your own infrastructure
| Feature | Basic | Pro | Premium | Intelligence |
|---|---|---|---|---|
| Deployment | ||||
| Runs on your own servers | ||||
| Your own database instance | ||||
| Patient data stays on site | ||||
| Docker Compose install | ||||
| High-availability Postgres | ||||
| Fully air-gapped operation | ||||
| Clinical modules | ||||
| Patient records & MRN | ||||
| OPD & IPD management | ||||
| Billing (GST 2.0) | ||||
| Pharmacy | ||||
| TPA / Insurance claims | ||||
| OT scheduling | ||||
| Lab (LIS) | ||||
| Radiology (RIS) | ||||
| Blood bank & CSSD | ||||
| HR & Payroll | ||||
| Procurement & inventory | ||||
| Multi-branch management | ||||
| Custom clinical workflows | ||||
| ✦ Asha AI — Intelligence plan | ||||
| WhatsApp patient agent | ||||
| TPA denial risk predictor | ||||
| Discharge summary generator | ||||
| Clinical decision support | ||||
| Demand & operational forecasting | ||||
| On-prem model (no cloud calls) | ||||
| ABDM compliance | ||||
| M1 — ABHA creation & linking | ||||
| M2 — Digital prescriptions | ||||
| M3 — Health Information Exchange | ||||
| FHIR R4 export / import | ||||
| Compliance & security | ||||
| NABH statutory registers | ||||
| DPDP Act 2023 safeguards | ||||
| Role-based access control | ||||
| Audit log retention | 1 yr | 3 yr | 7 yr | 7 yr |
| BAA / DPA agreement | ||||
| PCPNDT Form-F automation | ||||
| Support | ||||
| Email support | ||||
| WhatsApp support | ||||
| Phone support | ||||
| SLA | 48 h | 8 h | 2 h | 1 h |
| Dedicated account mgr | ||||
| On-site installation | ||||
How an install works
Nadi ships as a Docker Compose stack. It runs on a single Linux server for smaller sites, or across a cluster with high-availability Postgres for larger ones. Nothing depends on our cloud.
- 1
Scoping call
We size the deployment against your beds, users, departments and existing systems.
- 2
Install
Docker Compose stack on your servers. Premium and Intelligence include on-site installation.
- 3
Migration
We import patients, billing history and stock from your current HMS.
- 4
Handover
Staff training, runbook, and a support line with your plan's SLA.
Common questions
Why is there no price on this page?
Because there is no single number that would be honest. Every deployment runs on your hardware, imports data from a different legacy system, and enables a different set of modules. We scope it with you and quote once, rather than publishing a per-seat rate that nobody actually pays.
Is this a subscription?
No. Nadi is licensed for on-premise deployment, not sold as a per-seat monthly cloud service. Commercial terms — licence, support period and update cadence — are agreed in the contract.
What is the difference between the four plans?
The modules that are enabled, the scale the deployment is built for, and the support SLA. The software is the same on-premise product in all four.
Where does our patient data live?
On your servers, in your own PostgreSQL instance, inside your network. There is no shared multi-tenant database and no replication to our infrastructure.
Can Nadi run fully air-gapped?
Yes, on the Intelligence plan. Asha runs against a local model (Ollama) on your own GPU instead of a cloud AI provider, so no clinical text leaves the building. ABDM features need outbound connectivity to the national gateway, so those are the exception.
What hardware do we need?
It varies, and honestly we can't tell you before we look. Compute depends on patient volume, how many staff are on the system at once, whether you run imaging, and how long you retain records — a 40-bed clinic and a 40-bed day-surgery unit need very different machines. Basic and Pro typically sit on a single server; Premium scales out; Intelligence needs a GPU if you want Asha running entirely on-prem. We size it properly during the scoping call rather than publish a number that might not hold for you.
Can we move up a plan later?
Yes. Higher plans unlock additional modules on the same installation, and depending on the load they add, may need more capacity underneath. Your data stays where it is — there is no migration between plans.
How is ABDM compliance handled?
All ABDM traffic goes through the NHA's official ABDM Wrapper, deployed alongside Nadi. Nadi never calls the national gateway directly and never stores ABHA tokens in plain text.
Talk to us
Let's scope your deployment.
Tell us your bed count and how you run today. We'll come back with a sizing and a quote.