Work›Client work›Dr. Gabi

Backend by Nautical Type · Mobile backend 2025 · Live on the App Store · Scaled

A medical advisor
in your pocket.

Dr. Gabi is an AI-native consumer health app that personalises advice to the individual using it — informed by the medical documents they upload, the products they scan, and the symptoms they describe. Nautical Labs built the entire backend: data model, APIs, AI workflows and integrations. The app is on the App Store under Dr. Gabi's own brand; the engine is ours.

View on App Store ↗ drgabi.com ↗

0 → 1000s

Users scaled

From private beta to public App Store launch

1M+

Foods scored, personalised

Against the individual's conditions and allergies

AI-native flows shipped

Health scores · document analysis · prescription drafting

SOC 2
+ HIPAA-aware

Privacy posture

Built without PII where possible

01 — Approach

A clear contract between the app team and us: they own the experience, we own the engine.

The engine, not the app.

Dr. Gabi came to us with the experience already designed and an app team building the iOS surface. What they needed was a backend that could handle three genuinely AI-native flows — personal health scoring, medical document analysis, and AI-assisted prescription requests — without ever feeling like a thin wrapper around a model.

We took the whole back of house: data model, APIs, auth, payments, AI orchestration, third-party integrations. The split made the brief small in any one place and crisp at the seams. The team shipped the App Store launch on time, and we kept iterating on the engine after launch as new flows landed.

Free from pharma influence. Personalised to the individual using it. The backend has to make that promise true on every request. — Dr. Gabi positioning

02 — Stack

Backend chosen for AI workflow ergonomics, not for novelty.

Built around the AI flows.

Three flows drove the architecture: a personal health score that updates as the user's history grows, document analysis that ingests lab work and prescriptions, and a prescription-request pipeline reviewed by clinicians. Each needed structured, auditable outputs — not free-form chat.

Backend

Node · TypeScript · structured API contracts per flow

Database

Postgres · row-level security · zero-PII where possible

AI

Tool-using LLMs · function calling · expert review loops

Document analysis

Multimodal extraction from labs, scripts, executive assessments

Privacy

SOC 2 · HIPAA-aware · audit trail on every clinical surface

Scale

Designed for the launch spike and the long tail after it

03 — Outcomes

Live on the App Store. The backend continues to evolve as new flows land.

What we built and shipped.

Personalised health scoring engine.
Combines vitals, lab work, lifestyle inputs and product scans into a single per-user score that meaningfully changes as new data lands.

Medical document analysis pipeline.
Ingests lab work, imaging reports, prescriptions, executive assessments and doctor's notes — turns them into structured context the rest of the app can use.

AI-assisted prescription request flow.
Users describe symptoms; the system drafts an appropriate request, surfaces contra-indications from their history, and routes to clinical review.

Product scoring against the individual.
The 1M-product catalogue scored not generically but against each user's conditions and allergies — the part that makes the app feel personal, every time.

Scaled 0 to thousands of users.
From private beta through the App Store launch and into steady-state. The backend has carried each phase without the team having to rewrite for the next one.

View on App Store ↗ drgabi.com ↗ Discuss a similar build

A medical advisor in your pocket.