Device-as-a-Service
Clinical AI
Recognised by DPIIT

The prescription pad is broken.
We're replacing it.

MetaboliQ replaces the paper prescription pad with a dedicated clinical device. Offline-first. Zero behaviour change. Every prescription becomes a structured, patient-owned, FHIR-aligned health record — automatically.

<2s
App ready · cold start
<100ms
Handwriting recognition
Day 14
Speed parity target
The Problem

Healthcare fails at the
moment of creation.

Every interoperability initiative fails at the prescription moment because doctors won't slow down. The hardware has to be faster than paper — or it gets ignored.

4–5Bn
prescriptions generated per year in India

All paper. No data.

Each consultation generates a prescription that disappears when the patient walks out. No structure. No searchability. The next doctor starts from zero.

8 errors
per handwritten Rx on average · Joshi et al., 2016 (549 Indian OPD prescriptions)

Illegible. Incomplete.

Missing dose: 14–53%. Missing duration: 31%. Illegible to pharmacist: 3–11.5%. These are documented norms, not outliers.

67%
patients have poor comprehension of their Rx · Das et al., 2025 (Indian J Pharmacol)

No continuity of care.

India has 101M diabetics and 315M hypertensives requiring longitudinal medication tracking. Paper cannot provide it. Every follow-up begins from scratch.

"The infrastructure exists. 86.64 crore ABHA accounts. The data creation layer does not."

The bottleneck is not technology. It is workflow.

01
Paper prescriptions break the data chain

India's 4–5Bn annual OPD prescriptions are handwritten. Patients leave with a paper slip. No structured data is created. No longitudinal record exists. The ABDM rails are built — but records still aren't flowing.

02
Existing software adds friction, not removes it

Every EMR or prescription app on the market requires doctors to change behaviour. They slow down the consult. They're heavier than paper. Adoption collapses at week 2. The bottleneck is workflow, not willingness.

03
No purpose-built clinical hardware exists for India

iPads are too expensive and iOS-locked. Generic Android tablets aren't optimised for clinic environments. No device exists that is purpose-built for the Indian prescription workflow — offline-first, EMR stylus, ABDM-aligned.

Our Approach

Device-as-a-Service.
Hardware is the moat.

MetaboliQ is a Device-as-a-Service company for Indian clinical infrastructure. We design, manufacture, and operate purpose-built clinical devices — starting with Glyph, the world's first dedicated smart prescription pad for Indian doctors.

Unlike software-only competitors, our hardware creates physical lock-in. A doctor who has our device on their desk doesn't switch — any more than a surgeon switches scalpel brands mid-procedure. The device is the habit. The habit is the moat.

Every prescription written on Glyph generates FHIR R4 structured data, feeds the patient's ABHA wallet, and trains our clinical AI — creating a compounding data asset that no competitor can replicate without years of prescription volume.

DaaS model Hardware lock-in FHIR R4 ABDM aligned Offline-first AOSP locked OS
V1 · Build Complete · Now
Glyph Standard
Kotlin + Compose app on Android tablet. Paper-replacement wedge. Proves clinical workflow before ODM hardware investment. V1 build complete on Tab S10 FE.
V1 build complete EMR stylus Offline-first FHIR R4 SMS delivery
V2 · Series A · ODM Device
Glyph E Ink
Purpose-built E Ink Android tablet via Onyx/Bigme ODM. Carta 1200 panel. Wacom EMR stylus. AOSP locked OS. Hardware encrypted. Leased at ₹2,999/month — zero upfront for clinic.
DaaS leased E Ink Carta 1200 AOSP locked IP52 clinic-safe MOQ 1,000
V3 · Series B · Category Creation
Glyph AI Booklet
Foldable booklet. Left panel: E Ink prescription surface. Right panel: LCD AI clinical co-pilot showing drug interactions, dosing flags, patient history from ABHA. Patent-protected dual-display architecture.
Patent pending Dual-display On-device AI Advisory only ₹5–15k/mo
How Glyph Works

Open. Write. Send.
Faster than paper.

The doctor writes as usual. The system structures, signs, delivers, and stores — invisibly, in under 10 seconds.

01 / CAPTURE

Doctor writes as usual

Handwritten on a digital pad — recognized on-device against an India brand-first drug vocabulary. Favourites, templates, "Repeat last Rx." Fully offline — no network dependency in the prescription flow.

Offline-first
02 / STRUCTURE

AI extracts and validates

NLP maps drug, dose, route, frequency, and duration to FHIR fields. Confidence scoring flags uncertain entries. High-risk medication alerts for doctor review.

FHIR R4 mapped
03 / DELIVER

Rx to patient in <10 seconds

Signed PDF generated on-device. SMS link sent before the patient reaches the door. Doctor sees the timestamp: "8.3 seconds." Faster than handing over paper.

SMS · Print
04 / STORE

Patient wallet updated

The prescription becomes a wallet entry — a running medication timeline, current meds list, and emergency pack the patient controls and shares with consent.

Patient-controlled
05 / COMPLY

FHIR R4 + ABDM by default

MedicationRequest structured to FHIR R4 and ABDM Implementation Guide. Immutable audit trail. Consent-gated sharing. DPDP-compliant from day one.

ABDM-aligned
06 / CONNECT

Ecosystem interoperability

Lab orders flow to connected labs. Diagnostic results return to the wallet. Pharmacy orders route automatically. Every connection multiplies the value of the record.

Labs · Pharma · Insurers
Core Technology

Three proprietary components.
Remove any one and it fails.

Each component is irreplaceable. Together they constitute a novel technical system with no commercial equivalent in India or globally.

Drug Vocabulary
Local · <100ms
NLP Pipeline
Confidence-scored
FHIR R4 Generator
On-device · ABDM IG
PDF + Delivery
<10s · SMS
Offline Sync
Zero loss · EventLog
Audit + Consent
DPDP · ABDM HIE-CM
Component A — Drug Vocabulary

2,000+ Indian OPD brand names mapped to WHO INN generics. Specialty-ranked prescribing frequency. Trigram fuzzy matching. 15MB SQLite FTS5 local database. Sub-100ms on-device handwriting recognition with zero network. Without it: reverts to generic-first vocabulary, product becomes slower than paper, abandoned in week one.

Component B — FHIR R4 Generator

On-device MedicationRequest generation conformant with ABDM Implementation Guide. Zero server dependency in prescription creation flow. iText 7 PDF generation on-device in <300ms. Without it: becomes an unstructured PDF tool indistinguishable from existing Indian e-prescription platforms. Data layer ceases to exist.

Component C — Confidence-Scored NLP

Clinical NLP trained on Indian OPD patterns. Per-field confidence scoring. High-risk medication categories (anticoagulants, insulin, paediatric dosing) trigger verification before commit. Provenance labelling per field. Without it: clinical deployment is medico-legally indefensible. Medication errors create liability for prescribing doctors.

Architecture

The data pipeline.

One prescription → signed PDF + patient wallet entry + FHIR R4 record. Invisibly. In under 10 seconds.

Zone 1 — Capture

Offline-first. Prescription creation never touches the network. Local event-sourced SQLite store. Works in power cuts, basements, zero-signal environments. Zero prescription loss.

Zone 2 — Structure

AI NLP extracts drug, dose, route, frequency, duration. Confidence scoring flags low-certainty fields. High-risk medication verification surfaced for doctor review. FHIR R4 MedicationRequest created in the background.

Zone 3 — Distribute

Patient wallet. Consent-gated sharing. ABDM HIE-CM integration. Lab results ingestion. Pharmacy orders. Insurer claims. Every connection raises switching cost for the entire ecosystem simultaneously.

Kotlin · ComposeSQLite FTS5NestJS · TypeScriptPostgreSQL + RedisHAPI FHIRCloudflare R2Google Cloud PlatformDPDP-compliant
For Doctors

Designed for 30–80 patients/day.
4–6 minute consult windows.

Speed first

Faster than paper
by day 14.

The only metric that matters for adoption. Engineering target: speed parity with paper by day 14 of pilot deployment. Pilot launching with seed funding.

Repeat last Rx covers 40–60% of OPD sessions

For chronic care patients, the previous prescription is the foundation. One tap pre-fills the entire prescription. Doctor reviews, adjusts doses, sends. Under 20 seconds total.

Glyph — Speed features

Repeat last Rx — one tap

Pre-fills previous prescription. Edit doses, adjust durations. No retyping for follow-up patients.

Specialty templates

T2DM review, HTN follow-up, URTI, paediatric fever. Common presentations in one tap.

Favourites ranking

Most-prescribed drugs surface first. Learned from the doctor's own prescription history.

Assistant mode

Receptionist enters patient demographics. Doctor only writes the prescription. No admin overhead in the consult.

Reliability

Offline-first.
Always.

Prescriptions are written locally first. Sync happens in the background when connectivity returns. Offline-first architecture validated in V1 QA: zero prescription loss across instrumented network-interruption tests on Tab S10 FE.

🔒

Zero prescription loss — guaranteed

Event-sourced local store with hash-chained audit log. Every write is append-only. Conflict resolution handles device replacement and mid-sync disconnection. Patient safety is non-negotiable.

Glyph — Reliability

Offline-first architecture

SQLite + Room ORM. All Rx written locally first. No UI blocking on network calls. Works in poor-connectivity clinics.

Background sync

WorkManager handles sync when connectivity returns. Doctor never waits for network. Delivery happens silently.

Immutable audit log

Every prescription write, edit, and delivery is hash-chained. Medico-legal traceability built in.

End-to-end delivery validated in V1 QA

Signed PDF on-device in <300ms; SMS hand-off in <10s via DLT-approved sender MTABLQ.

For Patients

Every prescription becomes
a permanent health record.

Patients receive their prescription by SMS in under 10 seconds — a link to the signed PDF from a DLT-approved sender. It is simultaneously stored in their ABHA wallet — structured, searchable, shareable with any doctor with consent. The longitudinal medication timeline builds automatically.

📱
SMS delivery in <10 seconds
Signed PDF link from DLT-approved sender (MTABLQ). Patient doesn't need to install anything. Works on any phone.
🔗
ABHA wallet — automatic
FHIR R4 MedicationRequest stored against patient's ABHA ID. Accessible to any ABDM-connected provider with consent.
🔒
Consent-gated sharing
DPDP-compliant. Time-bound. Purpose-bound. Revocable. Patient owns the data. Full audit trail of every access.
Ramesh K. · Health Wallet ABHA Linked ✓
Medication Timeline
💊
Metformin 500mg BD + Telmisartan 40mg OD
Dr. Venkataraman · 10 Apr 2026 · Diabetology · FHIR R4
🔬
HbA1c: 8.2% · FBG: 142 · PP: 196
Apollo Diagnostics · 5 Mar 2026 · Lab result
💊
Metformin 500mg BD · Telmisartan 40mg OD
Dr. Venkataraman · 12 Jan 2026 · Diabetology · FHIR R4
📋
Emergency pack — active medications
Available offline · shareable via QR · no network needed
Why MetaboliQ

The empty quadrant no one else owns.

Every structured product sits in the high-friction half. Every low-friction product produces no data. MetaboliQ is the only occupant of low friction + high structure.

ToolSpeed vs paperWorks offlineStructured dataSolo GP fitIndia brand vocab
Paper + WhatsApp photoFastestFullNonePerfectN/A
Practo RaySlower (weeks 1–4)PartialModerateWeakGeneric-first
Eka CareModerateLimitedGoodModerateMixed
HealthPlixModerateLimitedGood (specialty)PoorSpecialist only
MetaboliQDay 14 target ↑Full — core arch ↑FHIR R4 ↑Primary ICP ↑Brand-first ↑
Five compounding moats — a competitor must replicate all five

Hardware lock-in

A doctor with our physical device on their desk doesn't switch. DaaS leasing means MetaboliQ owns the hardware layer indefinitely. No software-only competitor can replicate this without years of ODM relationships.

Workflow lock-in

Doctor habit is the strongest moat in healthcare. Once "repeat last Rx" is muscle memory, switching means relearning 40+ patients per day. Churn cost is clinical, not financial — and clinical switching costs are total.

Data network effects

Every prescription enriches the dataset. Handwriting recognition improves. AI extraction sharpens. Safety signals calibrate for the Indian population. A larger dataset produces a better product — automatically.

Longitudinal health graph

After 12 months: medications, allergies, labs, referrals — all FHIR-mapped. No competitor can replicate this asset without replicating the years of structured capture that built it. The graph is irreplaceable.

Ecosystem integration layer

Every lab, hospital, pharmacy, and insurer that connects raises the switching cost for every other party simultaneously. MetaboliQ becomes load-bearing infrastructure — not replaceable by a single competing app.

Vision

"Stripe didn't build a better payment form. It became the infrastructure. MetaboliQ doesn't build a better clinic management system."

Not another EMR.
The default interface for
healthcare data creation.

Every prescription a doctor writes on MetaboliQ becomes a permanent, portable, patient-owned health record — automatically, without the doctor thinking about data infrastructure for a single second. The data creates itself.

Stripe for payments
became infrastructure
UPI for transactions
became the standard
MetaboliQ for health data
becomes the default
Product Roadmap

Prescription pad today.
Health infrastructure tomorrow.

Three phases. Each self-funding the next. Each building on the moat of the last.

V1
2026
Q2–Q3

Paper Replacement

Offline-first handwritten prescription. On-device PDF. SMS delivery in <10s. India brand-first drug cache. Repeat last Rx. Favourite templates. Assistant mode. Launch pilot in Hyderabad → paid conversion begins.

Offline-firstSMS deliveryHandwriting recognitionFHIR-ready arch
V2
2026
Q3–Q4

Patient Wallet + Lab Integration

Patient medication timeline and clinician viewer. Consent-based sharing with audit logs. Emergency pack QR. Handwriting mode beta. First lab integration — structured order flow, zero call-backs.

Patient timelineClinician viewerEmergency QRLab integration
V3
2027+

Interoperability + API Platform

ABDM HIE-CM consent exchange. ABHA linking. AI medication reconciliation. Lab result ingestion. DICOM imaging. Full FHIR R4 + ABDM IG conformance. National health data infrastructure.

ABDM HIE-CMABHA linkingAI reconciliationAPI marketplace
Join the Pilot

Ready to replace
the prescription pad?

Recruiting 30 doctors in Hyderabad for our launch pilot — GP, paediatrics, diabetology. Onboarding begins post-seed close.

tushar.langer@metaboliq.in · metaboliq.in · Hyderabad