ApneDoctors

Swipe Left on Fake Advice. Match with Real Doctors, Backed by AI

Comment

Inspiration

In a country of 1.4 billion, where Google is the first doctor and WhatsApp forwards are the second, we asked:

“Why the hell isn’t instant, verified, AI-powered healthcare already a thing in India?”

That one question launched ApneDoctors — a platform born not in some boardroom, but out of real frustration, personal pain, and a dream to put real doctors + real-time AI in everyone’s pocket.

Whether it’s a college student at 2AM with a panic attack or a farmer in a remote village with no access to clinics, ApneDoctors delivers help, not excuses.

What it does

ApneDoctors is your 24x7 AI-powered medical lifeline, combining instant chatbot diagnosis, real doctor consultations, and verified health tracking — all in one blazing-fast, beautifully designed app.

AI Symptom Checker — Trained on real medical books, not Reddit threads.

Verified Doctor Onboarding — Licensed only. No frauds. No quacks.

End-to-End Health Privacy — HIPAA-grade encryption because your health ain’t for sale.

Multilingual Support — English, Hindi, Hinglish, and regional languages (coming soon).

Seamless App Experience — For rural India or urban Gen-Z, it just works.

How we built it

We went full-stack savage on this:

Frontend: Flutter (mobile) + React (web) with Tailwind for that slick Gen-Z UX.

Backend: FastAPI with Python, async like a beast.

LLM: Hugging Face Models fine-tuned for symptom mapping, wrapped in LangChain.

Doctor Onboarding: Aadhaar + DigiLocker + NMC real-time scraping for verification.

Symptom Checker: AI-powered triage + fallback rule-based model for edge cases.

Database: Firebase + PostgreSQL. Realtime, scalable, no BS.

Cloud: Deployed on GCP with Cloud Run + Firestore triggers + Load balancer.

Monitoring: Prometheus + Grafana with alerts on Slack.

Challenges we ran into

Verifying real doctors without manual KYC hell — we had to automate NMC scraping.

Building trust — Users don’t trust bots. So we added real-time human handover.

Latency — We pushed every millisecond using async APIs + vector caching.

AI hallucinations — We added grounding + retrieval augmentation using real medical textbooks.

Accomplishments that we're proud of

Built a working full-stack prototype in under 30 days.

Integrated real-time NMC verification for doctors — zero fake profiles.

AI chatbot with hallucination-guard using RAG + LangChain — safer than 90% of medical bots.

Simulated 679+ real-world patient cases to fine-tune triage and routing.

Scalable backend on GCP with sub-3-second response times.

Tested in edge locations with poor networks — optimized for Bharat.

Got early interest from local clinics and student health orgs.

Explainable AI integration with SHAP + fallback safety logic.

Live telehealth MVP working — doctors can take calls, send prescriptions, and chat securely.

What we learned

Doctors need verification. Not everyone with “Dr.” in their WhatsApp name is real.

Users need trust. We built layered transparency: license checks, reviews, AI cross-verification.

AI is powerful, but unsafe alone. So we trained it on trusted medical books + added real human review.

Speed wins. Every second counts. Our architecture is built for response under 3 seconds.

What's next for ApneDoctors

Prescription Generator – AI + doctor-written, smart formatting, pharmacy-ready.

Launch Beta in 2 Tier-2 cities + colleges with real patients and doctors.

Voice + WhatsApp Bot Support — Because not everyone types their symptoms.

Offline Mode + SMS Support for remote/rural areas.

AI Health Timeline — A dynamic chart of your health over time, like Spotify Wrapped for your body.

Family Health Tracker — One app, one login, full family care.

Insurance + eHealth Card Integration — Connect health records with policy coverage.

Government/NGO Partnerships — For large-scale public health deployments.

Built With

Share this project:

Updates