Case study · In-house AI productOur own product

Building and Running Clinizy Care: An AI-Powered, Multi-Tenant Hospital SaaS for Indian Clinics

Clinizy Care is the hospital management software Brynex Labs builds and runs for India's clinics, nursing homes and small hospitals. We own it end to end: eleven modules on one patient record, 24 built-in automations, WhatsApp built in, AI clinical notes in early access, and a full interface in English, Hindi and Hinglish.

Published Sep 30, 2026Updated Oct 4, 20267 min read

Meet the AI agents
Illustration: Clinizy Care at the centre of six AI agents: Bol, Saathi, Awaz, Nazar, Buddhi and Setu.

Key takeaways

What it is
AI-powered hospital management software for India's clinics, nursing homes and small hospitals, built and run by Brynex Labs.
The AI
Six agents cover documentation, the front desk, outreach, the owner's brief, the pharmacy and analytics. Bol is in early access; the other five are on the roadmap.
The core
Eleven modules on one patient record, 24 live automations and a fail-closed multi-tenant guard.
For you
The same stack and playbook we use for clients, proven first on our own product.

Results

  • 11

    Modules

    on one shared patient record

  • 24

    Built-in automations

    running on their own, 24/7

  • 6

    AI agents

    Bol in early access, five on the roadmap

  • 3

    Interface languages

    English, Hindi & Hinglish

Six AI agents, one clinic

Clinizy Care is AI-first by design. Six agents cover the clinic end to end: documentation, the front desk, patient outreach, the owner's morning brief, the pharmacy and analytics. Bol is in early access today. Saathi, Awaz, Nazar, Buddhi and Setu are on our roadmap, and we label them that way. Under all six sits a deterministic core and 24 automations that already run in production.

  • AI agentEarly access

    बोलBol

    AI clinical notes

    The doctor dictates the way they actually talk, and Bol drafts a structured clinical note to review, with a guardrail that rejects non-clinical input. Consult time goes to the patient, not the keyboard.

    Safeguard: Nothing is finalised: the doctor reviews every draft, and a guardrail rejects non-clinical input.

  • AI agentRoadmap

    साथीSaathi

    WhatsApp front desk

    An AI agent on the clinic's own WhatsApp number that books and reschedules, shares the queue position, sends reports and payment links, and hands anything clinical to a person.

    Safeguard: A small, bounded toolset with nothing clinical. Anything that sounds like a symptom goes to a person with the chat attached.

  • AI agentRoadmap

    आवाज़Awaz

    Reminder calls

    A 45-second call in the patient's language for a due follow-up, a refill or a ready report. It takes one spoken answer (yes, no or a day) and records it.

    Safeguard: One short script and one spoken answer. It records the reply and never tries to hold a conversation.

  • AI agentRoadmap

    नज़रNazar

    The 8 am owner brief

    Every morning the owner gets a WhatsApp brief: unbilled lab orders, missed follow-ups, stock about to expire. Every figure comes from a database query; the AI only ranks and explains.

    Safeguard: Every figure comes from a database query. The model only ranks and explains, so it cannot invent a number.

  • AI agentRoadmap

    बुद्धिBuddhi

    Pharmacy forecasting

    Forecasts demand from what the clinic's doctors actually prescribe, drafts the purchase order, and plans how to clear near-expiry stock while the supplier return window is still open.

    Safeguard: It drafts the purchase order and the rescue plan. The pharmacist edits and approves.

  • AI agentRoadmap

    सेतुSetu

    Ask your clinic

    Owners ask in plain Hindi or Hinglish and get a chart back. The AI only chooses from pre-approved, clinic-scoped reports, so it can never write its own queries or see another clinic's data.

    Safeguard: It picks from pre-approved, clinic-scoped reports and fills in the parameters. It never writes its own queries.

The product, in production

A real screen from Clinizy Care, shown with demo data.

Clinizy Care owner dashboard for a demo clinic, showing 30-day revenue, OPD visits, bed occupancy and pharmacy sales with trend charts

How it fits together

One patient record sits at the centre. The agents read and propose on top of it, and everything that leaves the building goes through queues.

  1. Experience

    What clinics touch

    • Clinic web app in the browser
    • Owner dashboard on a phone
    • English, Hindi and Hinglish
    • Public booking
    • Waiting-room display
  2. AI agents

    Read, draft and propose

    • Bol
    • Saathi
    • Awaz
    • Nazar
    • Buddhi
    • Setu
  3. Core services

    The deterministic heart

    • 11 modules, one patient record
    • 24 automations
    • GST billing engine
    • Role-based access
    • Audit trail
    • Fail-closed tenant guard
  4. Delivery

    Everything that leaves the building

    • Queues with dead-letter handling
    • WhatsApp Business API
    • Payments
    • PDF and report workers
  5. Data and cloud

    Where it runs

    • MongoDB
    • AWS, Mumbai region
    • Encrypted in transit
    • Automated tests in CI
On this page

About Clinizy Care

Clinizy Care is hospital management software for clinics, nursing homes and small hospitals across India. It covers the whole patient journey, from registration and the OPD token queue to prescriptions, pharmacy, lab, IPD, GST billing and WhatsApp follow-ups, on one shared patient record.

Unlike our other case studies, there is no client here. Clinizy Care is Brynex Labs' own product. We designed it, built it, and run it in production, and it is where every service we sell to clients gets tested first.

The Challenge

Most small clinics in India still run on paper registers, handwritten prescriptions and a pharmacy counter that bills separately. The software on the market was largely built for big hospitals or for other countries: priced out of reach, English-only, and unaware of GST or the way an Indian front desk actually works.

Building for this market meant hard constraints at once: a price a single-doctor clinic can start at (plans begin at ₹1,999 a month, excluding GST), an interface the front desk can use in Hindi or Hinglish, correct GST on every bill, and the reliability of one shared system that every clinic on it depends on at the same moment.

Our Approach

We treated Clinizy Care as a real business with real economics, not a demo. That shaped every architectural choice: one multi-tenant platform instead of an install per clinic, lean infrastructure in the AWS Mumbai region, and automation doing the repetitive work so a small clinic team never has to.

We also built it AI-native. Coding agents draft and senior engineers design, review and own every change, behind thousands of automated tests. It is the same way we build for clients, proven first on our own product.

What We Built

A complete, production hospital management platform across eleven modules, plus the automation, messaging and AI around it.

  • Front desk & OPD: patient registration, appointments and walk-ins in one live token queue, streamed to reception, the consulting room and a waiting-room display that shows initials only.
  • Clinical: digital prescriptions, patient records, lab orders and reports, and IPD with wards, beds, admissions, vitals and nursing medication charts.
  • Pharmacy & inventory: batch- and expiry-aware stock with purchase orders, supplier management and near-expiry alerts.
  • GST billing: CGST/SGST or IGST by place of supply, exempt consultations, MRP-inclusive pharmacy pricing, day-close reconciliation, UPI QR on invoices and a Tally export.
  • Autopilot: 24 built-in automations, from follow-up recalls and refill reminders to critical-result escalation and discharge workflows, with per-clinic controls.
  • WhatsApp: direct Meta Cloud API integration with approved English and Hindi templates, queued delivery, signed webhooks, quiet hours, opt-outs and quotas.
  • AI agents: six in total. Bol (clinical notes from dictation) is in early access. Saathi (a WhatsApp front desk), Awaz (reminder calls), Nazar (the 8 am owner brief), Buddhi (pharmacy forecasting) and Setu (ask your clinic in Hindi) are on the roadmap.
  • Growth engine: clinizy.in itself, with module, comparison and automation pages, a plain-language blog for clinic owners, and llms.txt for AI search.

The Engineering Decisions Behind It

Shared-collection multi-tenancy with a fail-closed guard: every clinic lives in the same database, and a tenant filter is injected into every query, count, aggregate and write. With no tenant in context, queries return nothing. An integration suite against a real database proves one clinic can never read another's data.

Queues between the product and the outside world: WhatsApp, email, PDF and report jobs run through AWS queues with dead-letter queues and an idempotent worker, so a slow third party never slows down the front desk.

Money as integers, time in IST: amounts are stored in paise end to end, and financial years, day-close and bill numbering are computed explicitly in Indian Standard Time, which removes a whole class of rounding and midnight bugs.

Three languages with a hard gate: English, Hindi and a separately written Hinglish locale, with a translation-parity test that fails the build if any screen is missing a string.

Access control built in: seven staff roles with one permission matrix shared by API and interface, two-factor sign-in and an audit trail of every change, built to be DPDP Act 2023 aligned.

The Impact

Clinizy Care is live at clinizy.in: eleven modules, 24 automations and a three-language interface, sold on public pricing with a 30-day free trial. For Brynex Labs it is also a working proof of everything we sell: multi-tenant SaaS engineering, AI agents with guardrails, workflow automation and SEO, all running in our own production.

That is the practical difference for clients. When we recommend an architecture, an AI pattern or an SEO plan, we are not guessing. We have already shipped it for ourselves, and we live with the results every day.

How we worked

  1. 1

    Start from the front desk

    We started with the people who live in the software: owners, doctors and receptionists, and where their minutes actually go.

  2. 2

    Build the deterministic core first

    One patient record, eleven modules, correct GST and a fail-closed tenant guard. AI gets nothing to stand on until the data underneath is right.

  3. 3

    Add AI where the minutes are

    Documentation first. Each agent gets a bounded toolset, a guardrail and a human review step before it touches a patient or a rupee.

  4. 4

    Run it ourselves

    We operate Clinizy Care in production, so every agent is judged against real clinic days rather than a demo.

  5. 5

    Ship, measure, extend

    Weekly releases, thousands of automated tests as a safety net, and a roadmap that says plainly what is live and what is next.

Technology

  • React
  • Vite
  • Tailwind CSS
  • Node.js
  • Express
  • MongoDB
  • Mongoose
  • Zod
  • i18next
  • AWS (EC2, S3, SQS, SES)
  • Meta WhatsApp Cloud API
  • Cashfree
  • Jest & Vitest
  • GitHub Actions
  • Sentry

Lessons you can reuse

  • Let the database decide

    Agents choose, draft and explain. Figures and records come from queries, so a model can never invent a number.

  • Fail closed, then prove it

    Our first tenant guard covered reads but not every write path. We extended it, then proved isolation with tests against a real database.

  • Keep a person in the loop

    Every agent drafts or proposes. A doctor, pharmacist or owner confirms before anything is signed, ordered or sent.

  • Say what is live

    We label every agent Early access or Roadmap on our own site. Buyers trust a product that is honest about its edges.

Questions

Is the AI in Clinizy Care live?

Bol, the clinical-notes agent, is in early access. Saathi, Awaz, Nazar, Buddhi and Setu are on our roadmap and not yet available. The 24 workflow automations are live today.

Who builds and runs Clinizy Care?

Brynex Labs designed it, builds it and operates it. clinizy.in is the home for the product, its features and its pricing.

Where is patient data hosted?

Clinizy Care is hosted on AWS in Mumbai, encrypted in transit, and built to be DPDP Act 2023 aligned. Compliance remains the clinic's responsibility as data fiduciary.

Can Brynex Labs build something like this for us?

Yes. The same team builds multi-tenant SaaS, AI agents and WhatsApp automation for clients, and healthcare is our proven vertical.

Your AI technology partner

Building with AI agents? Bring us the problem. We design, build and run it with you, and we can show you what we already run ourselves.

  • Agents built for production

    Evals, guardrails and observability ship with every agent, so it keeps working after launch.

  • A team that runs a live AI product

    Clinizy Care is ours. We know what breaks in production because we carry the pager.

  • Full-stack ownership

    App, API, AI and cloud from one senior team, with no hand-offs between vendors.

  • You own everything

    100% code and IP ownership, fixed-scope milestones and weekly demos.

Have an AI product to build?

Tell us what you're building. You'll talk to the engineers who ship AI agents for a living, and get a fixed-scope proposal before you commit.

All case studies