Case studies / AI Health Assistant

Client success story AI Engineering Digital Health

AI Health Assistant brings 24/7 AI health guidance to patients

From a symptom typed at 3am to a lab report explained in plain language and a follow-up check-in, all in one conversation. Built end to end across web, Android and iOS.

AI Health Assistant

HealthTech · AI virtual health assistant


24/7

AI guidance from symptom to next step

NetSet Software AI product engineering partner

At a glance

128k+

Patients Helped

5s

Median First AI Response

22

Languages Supported

6+

AI Specialist Assistants

The partnership

AI Health Assistant set out to shorten the gap between a health worry and a useful answer. A patient might wait three weeks for a specialist, or spend two weeks searching symptoms online. Clinics lose time to intake forms, repeat questions and missed follow-ups.

NetSet Software designed and built the AI healthcare product end to end: the patient web portal, Android and iOS applications, real-time AI chat, user accounts, payment integration, and credit system, from product design to launch.

Challenges we tackled

01 / Fast, responsible guidance

Answers had to arrive in seconds, in plain language, while making clear that the AI prepares users for a doctor and does not replace one.

02 / Unreadable medical documents

Lab reports arrive as PDFs and scans full of values, ranges and arrows. Users needed to know which numbers changed and what to do next.

03 / Two very different users

Patients want quick, clear answers. Doctors need a public profile and an AI receptionist that handles routine chats for them.

04 / Multilingual, mobile-first

Most users in India reach the product on a phone and in their own language, including Hindi, Hinglish and regional languages.

05 / Sensitive data and payments

Health records, family profiles and payments demanded HIPAA-ready handling: encryption in transit and at rest, AES-256-GCM encrypted chat messages and access logging. Data is never sold or used to train public models.

06 / A sustainable business model

AI usage costs money per message, so the product needed a fair credit system that users could understand and track.

Our engineering approach

01 Approach

One conversation for the whole care journey

We designed the product around a single chat thread. Symptoms, report questions and follow-ups all happen in that thread, so users never fill the same form twice.

02 Approach

Real-time AI over WebSockets

Assistant replies stream to the web and mobile apps over dedicated sockets for general chat and for report Q&A. Users see the answer form as it is written.

03 Approach

Specialist assistants, not one generic bot

Cardiology, neurology, oncology, gynecology, diabetes care and pediatrics each have their own assistant, prompts and sample questions, so answers start from the right clinical context.

04 Approach

One product across web, Android and iOS

The Flutter apps mirror the web app’s features, brand, typeface and legal copy, and share the same backend, sockets and encryption format.

05 Approach

Localisation built in from day one

Every screen reads from translation catalogues in 22 languages, with automated tests that catch a missing key in any language.

06 Approach

Platform-native where it matters

Sign in with Google and Apple, biometric unlock, Apple Health and Google Health Connect, push notifications, in-app reviews and in-app updates all use each platform’s own machinery.

What we built

24/7 Virtual Health Assistant

Describe symptoms in plain words and get a structured reading and a clear next step in under a minute. Supports voice input.

AI specialist consultations

Dedicated assistants for cardiology, neurology, oncology, gynecology, diabetes care and pediatrics, each with guided sample questions.

Clinical Insights

Upload a lab report PDF or scan. Values are parsed in about 8 seconds, changes are flagged and explained plainly, and follow-up questions are answered.

Health 360

Ask "best neurologist near me" and get ranked doctors, hospitals and labs by rating, distance and reviews.

Family health profiles

Manage profiles, habits and medications for family members from one account.

Medical file vault

Folders for prescriptions, lab reports, medication files and visit documents.

Health tracking on mobile

Reads Apple Health and Google Health Connect on the device and charts 30-day trends. Data stays on the phone.

Credits and payments

Buy credit packs through Razorpay, track the credit cost of every AI chat, download invoices, or earn credits with rewarded ads on Android.

Technology and contribution

React.js Flutter Python WebSockets FastAPI Firebase Auth Firebase Cloud Messaging Razorpay Apple HealthKit Google Health Connect Meta Audience Network i18next AES-256-GCM LangChain OpenAI FAISS Hugging Face embeddings

NetSet Software owned product design, web and mobile engineering, the real-time AI chat layer, localisation, payments and store releases.

Three apps share one real-time AI backend
Web appReact · patients and families
Android appFlutter · Health Connect
iOS appFlutter · HealthKit
Python API and real-time AI layerAI chat, specialist assistants, Clinical Insights, credits
AI Modelchat and report analysis
Firebasesign-in and push notifications
Razorpaycredit pack payments

Web, Android and iOS talk to one backend, which streams AI replies back over WebSockets and connects to the AI model, Firebase and Razorpay.

How we delivered

6

Team members

A dedicated NetSet Software team

One team delivered design, web, mobile and the AI layer, so the product feels the same on every platform.

Discovery

Guide, never diagnose

Mapped the patient journey and set the rule that the AI guides and prepares but never diagnoses.

Web

React web platform

Launched the web portal with specialist assistants, Clinical Insights and credits.

22 languages

Localisation

Rolled out 22 languages across every screen, including Hinglish and regional Indian languages.

Mobile

Flutter apps for Android and iOS

Rebuilt the full experience in Flutter, then added mobile-only features: health tracking, biometric unlock and rewarded ads.

Ongoing

Continuous releases

Features ship in small releases to the web, Google Play and the App Store, with in-app update prompts.

Results and business value

01 Value

Answers in seconds, not weeks

Patients get a first AI response in a median of 5 seconds, at any hour, instead of waiting weeks for a specialist callback.

02 Value

Lab reports people can act on

Reports are parsed in about 8 seconds and explained in plain language, with a list of questions to take to the doctor.

03 Value

Less intake work for clinics

Automated intake and the AI Receptionist collect history and answer routine questions before the patient reaches the doctor.

04 Value

Reach across languages and devices

One product serves users on web, Android and iOS in 22 languages.

05 Value

A clear revenue model

Credit packs, transparent per-chat usage and rewarded ads cover AI costs while keeping entry free.

AI Health Assistant provides AI guidance, not a diagnosis. Users are always advised to consult a licensed doctor for medical decisions.

AI engineering

Bring AI into your healthcare product.

Planning an AI health app, a virtual assistant for your clinic, or a multilingual patient platform? NetSet Software can take it from idea to app store.

Discuss your AI project

Let's discuss your engineering plan

Share a few details. We'll get back to you shortly.

What would you like to discuss? *

Your details stay confidential · NDA available before detailed discussion