Why CareLoop is different
The idea, the ideology, and what we give that typical hospital projects don't.
The ideology
“Care should follow the patient — not the hospital's org chart.”
Almost every hospital management system is built from the hospital's point of view: one hospital, its departments, its billing counter. The patient's real journey, however, crosses many organisations — a symptom at home, a choice between hospitals, a queue, a prescription, a pharmacy, weeks of medicines, follow-ups, maybe a second opinion elsewhere. Each hand-off is where care breaks: wrong specialist, long waits, lost reports, unaffordable medicines, missed doses, no follow-up.
CareLoop is built from the patient's point of view as a care-continuity network: independent services (triage, hospitals, appointments, records, pharmacy, feedback, analytics) cooperate through standard APIs so the loop Symptom → Specialist → Visit → Medicines → Recovery → Feedback is never broken — and so any new hospital, pharmacy or service can plug in without rewriting the system.
1 · Patient-owned
Records travel with the patient across hospitals. The patient decides who sees what, for how long — and can see every access in an audit ledger.
2 · Network, not silo
Government, trust and private hospitals sit side by side with live wait, beds, prices and Trust Scores — so people choose with facts.
3 · Explainable intelligence
Every AI output — triage urgency, CareMatch ranking, sentiment, no-show risk, Health Score — shows the factors behind it.
Typical student hospital project vs CareLoop
What others don't give| Area | Typical hospital / appointment project | CareLoop |
|---|
What we give that others don't
Research gap → our answer
Following the reference structure: Industry Need → Research Gap → Novelty → Problem Statement → Solution → Architecture → Stack.
Architecture at a glance
Open live system view →Communication styles
Synchronous REST (request/response) between services through well-defined contracts, plus asynchronous publish/subscribe events through the Notification & Event Bus.
Data ownership
Database-per-service: each service owns its own data store and nobody reads another service's data directly — only through its API.
Resilience
Graceful degradation (e.g. hospitals still list if Feedback is down — just without Trust Scores), auto-restart of crashed services, timeouts and friendly 503s at the gateway.
SOA principles demonstrated
| Loose coupling | Services only know each other's API contracts; events are fire-and-forget. |
| Service reusability | Auth, Notification and Feedback/Sentiment are generic and reused by every other service. |
| Abstraction | The browser only sees the API Gateway; internal ports and data stores are hidden. |
| Autonomy | Each service is its own process with its own data; can be restarted or scaled alone. |
| Composability | Analytics and CareMatch compose data from 3–6 services into new capabilities. |
| Discoverability | Service registry with heartbeats + an API catalogue on the System page. |
| Statelessness | JWT identity is verified once at the gateway and propagated in headers. |
| Interoperability | JSON over HTTP, standard status codes, correlation IDs (x-request-id). |
Technology stack
- Runtime
- Node.js 18+ (each service = separate process)
- Services
- Express 5 REST APIs
- Gateway
- Custom reverse proxy · JWT · rate limiting · metrics · registry
- Security
- bcrypt password hashing, JWT (12h), role-based access, consent checks, internal service key
- Data
- JSON document store per service (swappable for MySQL/MongoDB)
- AI / analytics
- Rule-based triage engine, multi-criteria ranking, lexicon NLP sentiment, forecasting, risk scoring
- Integrity
- SHA-256 hash-chained audit ledger
- Frontend
- Multi-page HTML5 + CSS3 + vanilla JS, custom SVG charts (works offline)
- Integrations
- Redirects to Tata 1mg, Apollo Pharmacy, PharmEasy, Netmeds, Jan Aushadhi; tel:108; Google Maps directions
Problem statement (user story)
“I woke up with chest discomfort. I didn't know if I needed a cardiologist or a gastroenterologist, or whether it was an emergency. The nearest private hospital had a 2-hour queue; the government hospital was free but crowded; I had no way to compare. After the visit, my prescription cost ₹600 at the counter — later I found the same salts for ₹120. I forgot doses, my reports stayed in a paper file, and my next doctor at another hospital had no idea what the first one prescribed. I want one place that tells me where to go and how soon, gets me the best available doctor across hospitals, keeps my records with me, finds my medicines at the right price, reminds me to take them — and lets me decide who sees my data.”