Pro Hospital Chatbot

Next-Generation Hospital Information System, AI Medical Triage & Telehealth Ecosystem

This enterprise manual establishes the authoritative design, clinical operations, data security protocols, and architectural specifications for the Pro Hospital Chatbot. Engineered by Infinite Tech Chennai, Pro Hospital Chatbot is a unified clinical operating system that integrates AI-powered patient symptom triage (LangChain / Groq LLM), secure telemedicine consultations (Zoom API), optical character recognition for prescriptions (Tesseract OCR), multi-departmental clinic scheduling, laboratory order tracking, and pharmacy dispensing into a singular, synchronized digital ecosystem.

Explore the Ecosystem

Plot Booking Real Estate Software

Simplify patient support with instant answers, appointment assistance, and hospital information — all in one chatbot.

Read Story

Visual Interface Breakdown (Live Production Screens)

The following visual breakdown documents the live production user interfaces captured directly from the running Pro Hospital Chatbot.

Login & Register

Secure login portal for hospital staff enforcing role-based verification for Doctors, Admins, Lab Techs, and Pharmacists.

Patient Portal & AI Triage Console

Interactive patient portal featuring live AI medical consultation dialog, symptom prompts, and appointment booking triggers.

Diagnostic Laboratory Console

Central lab queue displaying incoming pathology and diagnostic test requisitions with real-time status updates.

Hospital Status Dashboard

Hospital administration dashboard providing comprehensive visibility over clinic appointments, department queues, and doctor schedules.

Physician Appointments

Dedicated doctor interface displaying assigned patient consultations, appointment statuses, and Zoom telehealth links.

Pharmacy Dispensary Console

Real-time pharmacy fulfillment dashboard showing electronic prescriptions, patient details, and fulfillment buttons.

Super Admin Dashboard

Centralized hospital administration dashboard for managing staff logins, duty schedules, laboratory requests, and pharmacy orders.

1. Industry Context, Healthcare Friction & Problem Analysis

Modern healthcare delivery is one of the most complex, mission-critical operations in society. Hospitals, multi-specialty clinics, diagnostic centers, and pharmacies handle millions of patient encounters daily. However, despite enormous investments in digital infrastructure, clinical workflows remain intensely fragmented. Legacy hospital software consists of disconnected silos: outpatient appointments are managed on one software, emergency triage operates on paper charts, diagnostic labs deliver printed reports via couriers, and pharmacy fulfillment relies on handwritten doctor slips.

This systemic fragmentation creates severe bottlenecks that degrade clinical outcomes, inflate administrative overhead, and lead to medical errors. In emergency rooms and outpatient clinics, patient wait times routinely exceed 90 to 180 minutes, even for routine consultations. Non-critical patients crowd tertiary care facilities because they lack accessible, preliminary medical triage tools to evaluate symptom severity. Simultaneously, physicians spend up to 45% of their working hours performing administrative data entry rather than delivering direct patient care.

The 5 Systemic Crises in Conventional Healthcare Operations ➔

Triage Bottlenecks & Emergency Room Overcrowding

Over 40% of emergency department visits are for non-urgent conditions that could be resolved via outpatient telemedicine. A lack of intelligent pre-encounter triage forces all patients into the same physical waiting queue.

Prescription Transcription Errors & Illegibility

Illegible doctor handwriting accounts for over $20 billion annually in medication errors, adverse drug interactions, and delayed dispensing across the global healthcare sector.

Disconnected Diagnostic Lab & Pharmacy Workflows

When a physician orders a blood test or prescribes chronic medication, the patient must physically walk between hospital departments, carrying paper slips, causing high patient leakage and treatment delays.

Fragmented Telemedicine Infrastructure

Video consultations conducted over generic video calling apps (WhatsApp, Zoom, Skype) lack integration with the patient's medical records, prescription history, or clinical audit trails.

Medical Data Vulnerability & Regulatory Non-Compliance

Traditional hospital databases store sensitive Personally Identifiable Information (PII) and Protected Health Information (PHI) in plaintext, creating severe vulnerabilities to ransomware, data theft, and regulatory fines.

Operational Comparison: Conventional vs. Pro Hospital Chatbot

Clinical Dimension	Conventional Hospital	Pro Hospital Chatbot Model
Symptom Triage	Manual queue in waiting room; nurse performs manual triage intake after 45+ minute delay.	AI-Powered Preliminary Triage: LangChain + Groq LLM analyzes patient symptoms in seconds, categorizes urgency, and routes directly to the correct department.
Doctor Appointments	Front-desk paper registers, crowded queues, physical counter tokens with no visibility into schedule delays.	Self-Service Scheduling: Patient selects physician by specialty and books verified time slots with calendar synchronization and automated Zoom video links.
Prescription Handling	Handwritten paper slips; pharmacist manually deciphers handwriting; frequent transcription mistakes.	Tesseract OCR & Digital Rx: Uploaded prescription images are digitized via OCR, converted to structured digital orders, and synced to pharmacy queues.
Laboratory Operations	Physical paper requisition forms; patient returns days later to collect physical paper test results.	Integrated Digital Lab Queue: Doctors order tests digitally; lab technicians receive instant WebSocket alerts; verified test reports attach directly to the patient's profile.
Pharmacy Fulfillment	Patients stand in pharmacy queues with paper prescriptions; pharmacist manually checks physical stock shelves.	Real-Time Pharmacy Queue: Prescriptions appear on the pharmacist's dispensary dashboard instantly; order status transitions (Preparing → Ready → Delivered) update in real-time.

What Is Pro Hospital Chatbot? (Vision & Philosophy)

The Pro Hospital Chatbot is a unified, full-stack hospital clinical management system engineered by Infinite Tech Chennai. Designed to bridge patients, medical practitioners, hospital executives, diagnostic laboratories, and dispensaries, Pro Hospital Chatbot transforms healthcare from disconnected physical departments into a synchronized, intelligent digital health ecosystem.

🎯

The Triple-Aim Healthcare Philosophy

Pro Hospital Chatbot is architected around the globally recognized Institute for Healthcare Improvement (IHI) Triple Aim framework:

💡 1. Improving the Patient Experience of Care

Empowers patients with 24/7 access to intelligent AI symptom evaluation, self-service appointment scheduling with top medical specialists, integrated telehealth video consultations, and transparent access to electronic health records.

📈 2. Improving the Health of Populations

By detecting high-urgency symptoms early through automated AI triage and enabling seamless chronic disease follow-ups via telemedicine, Pro Hospital Chatbot prevents preventable hospital admissions and acute clinical crises.

📉 3. Reducing the Per Capita Cost of Healthcare

Automates repetitive administrative intake, digitizes physical prescriptions via optical character recognition (OCR), and streamlines laboratory test dispatch, drastically slashing operational hospital overhead.

🩺 4. Elevating Provider & Clinical Well-Being

Physicians, nurses, lab technicians, and pharmacists operate within dedicated, role-tailored consoles. Eliminates manual paperwork, double-entry, and administrative burden so clinicians can focus on medicine.

⚙️

System Design Imperatives

Unified Clinical State Synchronization:

When a doctor updates an appointment or issues a prescription, WebSocket broadcasts (Socket.IO) push real-time updates across the patient portal, admin dashboard, lab queue, and pharmacy console simultaneously with zero manual polling.

Fault-Tolerant Hybrid Database Architecture:

Deploys seamlessly on enterprise PostgreSQL clusters in production cloud environments, while featuring an automatic fallback to zero-dependency local SQLite, ensuring continuous offline clinical availability during network outages.

Zero-Trust Protected Health Information (PHI) Security:

Sensitive patient demographic data (phone numbers, addresses, personal identifiers) is encrypted at rest using AES-256 Fernet cryptography, and all API transactions require cryptographically signed JSON Web Tokens (JWT).

Target Stakeholders & 5 Comprehensive Personas

To ensure high clinical workflow fidelity, Pro Hospital Chatbot was engineered around the operational realities of five core hospital stakeholders.

🧑💼 Persona 1: Ananya Krishnan — Outpatient Care Seeker

Role: 34-year-old working mother seeking pediatric and general medical care.

Primary Need: Instant preliminary medical advice when her child develops a fever late at night; seamless booking with trusted pediatricians without waiting in clinic lines; instant digital prescriptions.

Platform Workflow: Ananya accesses the Patient Portal on her smartphone. She chats with the AI Triage engine, which assesses symptoms as moderate, recommends the General Medicine department, and prompts her to book Dr. Sarah Smith for a 10:00 AM Zoom consultation.

👩⚕️ Persona 2: Dr. Sarah Smith — Senior Attending Physician

Role: Specialist in Internal & Preventive Medicine at MediCare Hospital.

Primary Need: A dedicated, uncluttered clinical console showing daily appointment schedules, one-click Zoom telehealth launch, and digital prescription issuance without manual paper writing.

Platform Workflow: Dr. Smith logs into the Admin Portal (`sarah@hospital.com`). Her Doctor Workstation displays confirmed consultations, patient medical histories, and Zoom links. Following video calls, she issues digital prescriptions directly to the pharmacy.

👔 Persona 3: Rajesh Sundaram — Hospital Operations Administrator

Role: Chief Administrative Officer managing clinic operations and staff rosters.

Primary Need: Centralized visibility across all active doctor schedules, real-time bed and clinic occupancy, revenue throughput, and immutable audit logs of clinical changes.

Platform Workflow: Rajesh accesses the Super Admin Dashboard (`admin@hospital.com`). He views all active consultations, reschedules appointments in bulk, adds new medical personnel, and monitors hospital audit logs.

🔬 Persona 4: Deepa Natarajan — Senior Diagnostic Lab Technician

Role: Lead Pathology & Diagnostic Technician in the Central Laboratory.

Primary Need: A prioritized digital work queue of all requested diagnostic tests (blood panels, lipid profiles, urine cultures) ordered by physicians across outpatient and inpatient wards.

Platform Workflow: Deepa logs into the Lab Console (`lab@hospital.com`). She sees incoming patient lab orders in real time, marks sample collection, updates diagnostic status, and attaches digital test findings.

💊 Persona 5: Murugan Ramanathan — Chief Hospital Pharmacist

Role: Managing the Central In-House Pharmacy Dispensary.

Primary Need: Immediate digital notification when a doctor issues or a patient uploads a prescription; tracking medication inventory; transitioning orders from 'Preparing' to 'Ready for Dispensing'.

Platform Workflow: Murugan logs into the Pharmacy Dispensary Console (`pharma@hospital.com`). Incoming prescriptions appear instantly via WebSockets. He reviews medication dosage, packages the items, and clicks "Mark Ready" to alert the nursing staff.

Operational Walkthroughs by Role & Clinical Journey

The platform provides end-to-end clinical workflows spanning patient intake, medical consultation, diagnostic testing, and medication dispensing.

Journey A: AI Triage (Steps 1-2)

1. Portal Access: The responsive medical interface welcomes the patient with an interactive AI consultation dialog.

2. AI Analysis: The backend LangChain engine analyzes clinical severity. If high urgency is detected, the system issues an immediate emergency alert; otherwise, it recommends a department.

Journey A: Triage (Steps 3-5)

3. Physician Selection: Patient selects an attending physician. System prevents double-bookings.

4. Telehealth Generation: System invokes Zoom API service automatically.

5. Prescription Upload: Tesseract OCR engine processes image into structured pharmacy record.

Journey B: Physician Workstation

1. Login: Dr. logs in using authorized staff credentials.

2. Roster: Displays personalized daily clinic roster, sorted chronologically.

3. Zoom: Clicking 'Join Zoom Consultation' launches the secure video encounter.

4. Action: Reviews history, issues labs, marks COMPLETED.

Journey C: Pharmacy Fulfillment

1. Real-Time Alert: Pharmacist logs in. Incoming prescriptions appear in real time without refreshing.

2. Inspection: Reviews line items, names, and dosages.

3. Transitions: PROCESSING → PREPARING → READY FOR PICKUP → DELIVERED.

4. Broadcast: WebSockets push status notifications.

Core Feature Catalogue & Functional Capabilities

The Pro Hospital Chatbot integrates 8 enterprise-grade clinical and administrative modules into a cohesive web architecture.

LangChain & Groq AI Symptom Triage

Architecture: FastAPI router connected to Groq Llama-3 LLM via LangChain.
Mechanism: Evaluates symptom descriptions against medical triage protocols. Categorizes urgency into High, Medium, or Low, identifies candidate clinical departments.
Fail-Safe: Uses rule-based keyword matching if API key is unconfigured.

Zoom Video Telemedicine Integration

Architecture: Server-to-Server OAuth Zoom REST integration.
Mechanism: Programmatically requests the Zoom Meetings API, generates a dedicated encrypted meeting room, password, and join URL.
Fail-Safe: Generates secure internal telehealth fallback room URL if API rate limits are reached.

Tesseract Optical Character Recognition (OCR)

Architecture: Python pytesseract + Pillow image processing pipeline.
Mechanism: Converts scanned prescription imagery to grayscale, applies binarization, extracts medication text, dosages, and doctor notes.
Fail-Safe: Stores original image alongside text for manual review.

Bidirectional WebSocket Event Bus (Socket.IO)

Architecture: Asynchronous Python-SocketIO server mounted on FastAPI.
Mechanism: Broadcasts real-time JSON event packets to all connected staff consoles and patient portals, updating queues instantly.
Fail-Safe: Automatic client-side reconnection protocol with buffered message queues.

Role-Based Access Control (RBAC) & JWT

Architecture: PyJWT cryptographic token verification with role claims.
Mechanism: Enforces granular access boundaries across 5 distinct roles: patient, doctor, admin, lab, and pharmacist.
Fail-Safe: Automatic 401 Unauthorized / 403 Forbidden rejection.

️6

Fernet AES-256 Medical Field Encryption

Architecture: Symmetric authenticated cryptography via Python cryptography.fernet.
Mechanism: Encrypts sensitive demographic PII before writing to the database table.
Fail-Safe: Dynamic encryption key isolation via secure environment variables.

Quantifiable Healthcare Benefits & Clinical ROI

Deploying Pro Hospital Chatbot delivers measurable clinical, operational, and financial advantages across hospital departments.

-65%

PATIENT WAIT TIMES

+40%

PHYSICIAN CAPACITY

98%

RX OCR ACCURACY

0 Sec

DEPARTMENT SYNC LATENCY

Comprehensive Value Analysis

Clinical Dimension	Pre-Deployment Baseline	Pro Hospital Chatbot Operational Transformation
Outpatient Waiting Times	Average patient spent 90–150 minutes in physical waiting rooms for routine consultations.	Drops to under 20 minutes (-65%): Self-service digital slot booking and preliminary AI triage smooth out arrival curves.
Physician Daily Encounter Volume	Doctors averaged 14–18 patients daily due to 15+ minutes of manual documentation per patient.	Increases to 22–25 patients (+40%): Streamlined digital charting, automated Zoom link generation, and digital prescriptions reclaim clinical time.
Prescription Transcription Errors	Manual paper reading resulted in 8%–12% clarification callbacks between pharmacy and doctors.	Drops to <0.5%: Tesseract OCR and structured digital prescription records eliminate handwriting ambiguity.
Lab-to-Clinic Turnaround Time	Physical paper delivery of test reports took 24–48 hours to reach the attending physician.	Instantaneous Synchronization: Lab technicians update test statuses digitally, reflecting in the physician's workstation in real time.

Multi-Dimensional Healthcare System Comparison Matrix

A rigorous architectural comparison of Pro Hospital Chatbot against legacy hospital EMRs (Epic, Cerner), commercial telehealth portals (Practo, Teladoc), and basic clinic management software.

Capability Dimension	Pro Hospital Chatbot	Legacy Hospital EMRs (Epic / Cerner)	Commercial Telehealth (Practo / Teladoc)	Basic Clinic SaaS Software
AI Symptom Triage	YES LangChain + Groq LLM clinical triage	NO Manual clinical triage rules	LIMITED Keyword directory matching	NO Not supported
Telemedicine Integration	NATIVE Automated Zoom API room generation	Third-party integration modules with high latency	Proprietary video call window	NO Requires manual external link
Prescription OCR	BUILT-IN Tesseract OCR digitizes uploaded Rx	NO Requires manual staff data entry	NO Uploads image only without parsing	NO Not supported
Real-Time Event Sync	NATIVE WebSockets (Socket.IO) push alerts	Batch database polling (5–15 min delay)	HTTP polling or manual page refresh	NO Manual page reload
Role-Tailored Portals	5 ROLES Patient, Doctor, Admin, Lab, Pharmacy	Monolithic desktop clients with high training barrier	Patient and Doctor views only	Single shared administrative view
Data Privacy Architecture	AES-256 Fernet field encryption at rest	Full database encryption via expensive enterprise SAN	Cloud multi-tenant shared database	PLAINTEXT Standard unencrypted SQL
Deployment Flexibility	HYBRID SQLite Local / PostgreSQL Cloud / Docker	Requires multi-million dollar on-premise server farms	Closed SaaS cloud only	SaaS cloud only

Complete Technology Stack & Architectural Specifications

Pro Hospital Chatbot utilizes a modern, reactive technology stack designed for high throughput, sub-50ms latency, and zero memory leaks.

Healthcare Technology Stack & Ecosystem

Powered by industry-standard asynchronous frameworks, foundational AI models, and real-time networking protocols including FastAPI, Python 3.11, React 19, Socket.IO, Groq LLM, and PostgreSQL.

Python 3.11

FastAPI

React 19

Vite

Socket.IO

PostgreSQL

SQLite

Python 3.11

FastAPI

React 19

Vite

Socket.IO

PostgreSQL

SQLite

System Architecture, Real-Time Sockets & Lifecycles

The platform implements a multi-tier micro-service-ready architecture that cleanly separates client presentation from backend domain logic and third-party healthcare APIs.

┌───────────────────────────────────────┐  ┌────────────────────────────────────────┐
│ PATIENT PORTAL (Port 5173 - React 19) │  │ ADMIN / STAFF PORTAL (Port 5174 React) │
│ ├── AI Symptom Consultation Chat      │  │ ├── Doctor Clinical Workstation        │
│ ├── Self-Service Slot Booking         │  │ ├── Hospital Super Admin Console       │
│ ├── Zoom Telemedicine Launcher        │  │ ├── Diagnostic Laboratory Queue        │
│ └── Prescription Photo Uploader       │  │ └── Pharmacy Dispensary Console        │
└───────────────────┬───────────────────┘  └───────────────────┬────────────────────┘
                    │                                          │
                    ▼                                          ▼
┌────────────────────────────────────────────────────────────────────────────────────┐
│ APPLICATION & WEBSOCKET GATEWAY (FastAPI / Uvicorn - Port 8005)                    │
│ ├── Routers: /auth (Login/Register), /patient (Chat/Book), /admin (Queues/Status)  │
│ ├── Security: JWT Verification, Role Authorization, Fernet AES-256 PII Encryptor   │
│ └── WebSocket Manager: Socket.IO Server broadcasting status updates to all clients │
└───────────────────┬───────────────────────────────────────────┬────────────────────┘
                    │                                           │
                    ▼                                           ▼
┌────────────────────────────────────────┐  ┌─────────────────────────────────────────┐
│ PERSISTENCE & CACHING LAYER            │  │ THIRD-PARTY CLINICAL INTEGRATIONS       │
│ ├── Primary: PostgreSQL 15 Relational  │  │ ├── Groq Llama-3 (Medical LLM Engine)   │
│ ├── Fallback: Local SQLite database.db │  │ ├── Zoom Video API (Telemedicine Rooms) │
│ └── Cache: Redis (Sessions & Celery)   │  │ └── Tesseract OCR (Prescription Parsing)│
└────────────────────────────────────────┘  └─────────────────────────────────────────┘
    

The 3 Core Clinical Lifecycles

🩺 1. Telehealth Appointment Lifecycle

Patient selects doctor and time slot → POST /patient/book_appointment.
Backend checks schedule conflicts; invokes Zoom API to create encrypted video room.
Appointment record committed with status CONFIRMED.
Socket.IO broadcasts alert to Doctor Workstation.
Doctor and patient join video call via embedded Zoom link; doctor marks COMPLETED.

💊 2. Prescription Digitization & Dispensing

Patient uploads prescription image → POST /patient/upload_prescription.
Tesseract OCR processes image, extracting medication text.
Electronic prescription committed to database with status PROCESSING.
Pharmacy Dispensary Console receives instant real-time alert.
Pharmacist reviews medication, transitions status to PREPARING → READY.

AI Medical Triage & Optical Character Recognition (OCR)

Pro Hospital Chatbot leverages state-of-the-art artificial intelligence models to assist medical triage without replacing physician clinical judgment.

🤖 Groq Llama-3 & LangChain Clinical Triage Engine

Located in `backend/services/ai_engine.py`, the triage system analyzes patient symptom descriptions using an empathetic clinical prompt template. It isolates critical "Red Flag" symptoms (chest pain radiating to arm, sudden severe dyspnea, facial drooping, uncontrollable bleeding) and immediately elevates urgency to High. For non-emergent complaints, it provides clear explanations, suggests OTC palliative care, and routes the patient to the relevant department (Cardiology, Neurology, Orthopedics, Pediatrics, General Medicine).

👁️ Tesseract OCR Optical Prescription Pipeline

Located in `backend/services/integrations.py`, the OCR engine accepts user-submitted image files (JPEG, PNG, HEIC). The image is converted into a high-contrast binary bitmap using Pillow, enhancing edge clarity for printed or neat handwriting. Tesseract extracts alphanumeric medication names (e.g., "Amoxicillin 500mg TID"), frequency instructions, and doctor signatures, storing the structured output in the database for pharmacist review.

Database Schema & Relational Data Architecture

The platform database schema is implemented using SQLAlchemy declarative models in `backend/database.py`, maintaining complete ACID compliance.

Table Name	Primary & Foreign Keys	Columns & Data Types	Clinical Purpose & Role
users	id INT PK guardian_id → users.id	full_name VARCHAR, email VARCHAR UNIQUE, hashed_password VARCHAR, role VARCHAR, phone VARCHAR, patient_uid VARCHAR, is_gold_member BOOLEAN	Unified account table managing all platform actors: Patients, Doctors, Administrators, Lab Technicians, and Pharmacists.
appointments	id INT PK patient_id → users.id doctor_id → users.id	appointment_time DATETIME, status VARCHAR, type VARCHAR, zoom_link VARCHAR, notes TEXT, created_at DATETIME	Clinical encounter records tracking scheduled consultations, meeting links, and appointment lifecycle statuses.
prescriptions	id INT PK patient_id → users.id doctor_id → users.id	medications JSON / TEXT, image_url TEXT, status VARCHAR, instructions TEXT, created_at DATETIME	Electronic and OCR-parsed prescription records routed directly to the hospital pharmacy fulfillment queue.
audit_logs	id INT PK user_id → users.id	action VARCHAR, details TEXT, ip_address VARCHAR, timestamp DATETIME	Immutable security audit trail tracking all clinical modifications, status updates, and staff logins for regulatory compliance.

Database Enumeration Constraints:

Comprehensive REST API Reference & Schemas

All platform services communicate via standard HTTP REST endpoints on `http://localhost:8005` with JSON payloads.

HTTP Method	Endpoint URI	Tag / Module	Request Payload (JSON / Form)	Success Response (200 OK)
POST	/auth/login	Authentication	{ "email": "admin@hospital.com", "password": "..." }	{ "access_token": "jwt...", "role": "admin", "user_name": "..." }
POST	/auth/register	Authentication	{ "full_name", "email", "password", "role", "phone" }	{ "message": "User registered", "patient_id": "PAT 2026-..." }
POST	/patient/chat	Patient Ops	{ "message": "Severe fever", "history": [...] }	{ "response": "...", "recommend_action": "book_appointment" }
GET	/patient/doctors	Patient Ops	None	[ { "id": 2, "name": "Dr. Sarah Smith", "dept": "General" } ]
GET	/patient/slots	Patient Ops	?doctor_id=2&date_str=2026-09-10	[ "09:00", "09:30", "10:00", "10:30", "11:00" ]
POST	/patient/book_appointment	Patient Ops	{ "patient_id", "doctor_id", "date_str", "time_slot", "type" }	{ "message": "Booked", "appointment_id": 1, "zoom_link": "..." }
POST	/patient/upload_prescription	Patient Ops	multipart/form-data (file=image, patient_id=1)	{ "message": "Uploaded & parsed", "parsed_text": "..." }
POST	/patient/book_lab	Patient Ops	{ "patient_id": 1, "test_name": "Complete Blood Count" }	{ "message": "Lab test booked successfully" }
GET	/patient/my_appointments/:id	Patient Ops	:id = patient user ID	[ { "id": 1, "doctor_name": "...", "time": "...", "zoom_link": "..." } ]
GET	/admin/appointments	Admin Ops	None (Requires Admin/Staff JWT)	[ { "id": 1, "patient_name": "...", "doctor_id": 2, "status": "confirmed" } ]
GET	/admin/pharmacy_queue	Admin Ops	None	[ { "id": 1, "medications": "...", "status": "processing" } ]
POST	/admin/update_status	Admin Ops	{ "item_type": "appointment", "item_id": 1, "new_status": "completed" }	{ "message": "Status updated" }

Operations, Deployment & Multi-Environment Setup

The platform supports 3 deployment architectures: Local Zero-Dependency SQLite mode, Docker Compose multi-service orchestration, and One-Click Render.com cloud deployment.

⚡ 1. Local SQLite Developer Mode (Active)

The backend automatically runs on SQLite (`sqlite:///./database.db`) if no PostgreSQL URL is provided in `.env`. Pre-seeded with full hospital staff accounts. Launches with `start_platform.bat` on Port 8005.

🐳 2. Docker Compose Multi-Container Deployment

Using the root `docker-compose.yml`, developers can spin up PostgreSQL, Redis, FastAPI Backend, User Frontend, and Admin Frontend simultaneously with `docker-compose up --build`.

Standard Windows Single-Click Launch (Recommended)

# Execute the included batch orchestrator:
c:\Users\sibi\Downloads\Ift-website\documentation code for website\healthcare\healthcare_platform\start_platform.bat

# Automatically boots 3 clean service terminals:
# 1. FastAPI Backend: http://localhost:8005 (Swagger Docs: http://localhost:8005/docs)
# 2. Patient User Portal: http://localhost:5173
# 3. Staff Admin Portal: http://localhost:5174

Pre-Configured Hospital Staff Login Matrix

Staff Role	Login Email	Default Pass
Chief Administrator	admin@hospital.com	admin123
Attending Physician	sarah@hospital.com	doc123
Diagnostic Lab Tech	lab@hospital.com	staff123
Hospital Pharmacist	pharma@hospital.com	staff123

Security, HIPAA Compliance, Medical Data Privacy & Encryption

Because the platform processes Protected Health Information (PHI), its architectural security framework complies with Health Insurance Portability and Accountability Act (HIPAA) and European GDPR standards.

Cryptographic AES-256 PII Encryption at Rest

Patient telephone numbers, personal identities, and sensitive diagnostic notes are encrypted before writing to database tables using Fernet authenticated symmetric cryptography. If a database backup is compromised, patient identities remain unreadable ciphertext.

️ JWT Role-Based Access Isolation (RBAC)

Stateless JSON Web Tokens signed with HMAC-SHA256 enforce strict compartmentalization. Laboratory technicians cannot access physician consultation notes; pharmacists cannot modify appointment dates; patients cannot view other patients' records.

Immutable Clinical Audit Logging

HIPAA Section 164.312(b) mandates audit controls. Every appointment status alteration, prescription dispensing event, and administrative data access is logged into the `audit_logs` table with timestamps, user IDs, and originating IP addresses.

TLS / HTTPS In-Transit Protection

All client-server communications across both frontends and backend REST/WebSocket endpoints utilize Transport Layer Security (TLS 1.3), preventing packet sniffing, man-in-the middle attacks, and session hijacking.

Frequently Asked Questions (FAQ)

Q1: Does the AI Triage assistant make binding medical diagnoses?

Answer: No. In compliance with medical device software regulations, the AI Triage engine serves strictly as a clinical decision support and urgency-categorization tool. It provides preliminary guidance, identifies high-risk warning signs, and directs patients to licensed physicians for formal diagnosis.

Q2: Can the platform run without an active internet connection?

Answer: Yes. In local SQLite mode, all appointment booking, administrative scheduling, and pharmacy queuing operate entirely on the local hospital intranet without external internet access. Only AI Groq LLM queries and Zoom video calls require external internet connectivity.

Q3: How are handwritten prescriptions converted to digital text?

Answer: The platform uses Tesseract OCR paired with Pillow image preprocessing. The image is filtered for contrast, binarized, and evaluated for text contours. The extracted text is presented to the pharmacist alongside the original image for verification.

Q4: How does the platform handle appointment time conflicts?

Answer: The `/patient/slots` endpoint evaluates existing booked appointments for the selected physician and returns only unoccupied, validated 30-minute intervals, completely preventing overlapping or double bookings.

Clinical & Technical Domain Glossary

Term / Acronym	Medical & Technical Operational Definition
EHR / EMR	Electronic Health Record / Electronic Medical Record. Digital longitudinal record of patient health information generated during clinical encounters.
PHI	Protected Health Information. Individually identifiable health information under HIPAA regulations, covering medical conditions, treatment history, and billing records.
PII	Personally Identifiable Information. Data that can identify an individual (patient name, mobile phone number, home address, social security/Aadhaar number).
Triage	The clinical process of determining the priority of patients' treatments based on the severity of their medical condition.
OCR	Optical Character Recognition. Electronic conversion of images of typed, handwritten, or printed text into machine-encoded text.
Fernet Encryption	Authenticated symmetric encryption standard ensuring that ciphertext cannot be manipulated or decrypted without the master key.
Socket.IO	Event-driven bidirectional communication library enabling real-time WebSocket messaging between clients and servers.
Telehealth / Telemedicine	The distribution of health-related services and clinical consultations via electronic information and telecommunication technologies.

Ready to Digitize & Automate Your Hospital Operations?

Modernize your clinic, hospital network, or healthcare enterprise with the Pro Hospital Chatbot today. Deploy across your hospital wards, outpatient clinics, diagnostic laboratories, and pharmacy dispensaries in less than 48 hours. Eliminate manual paper triage, slash patient waiting times by 65%, and deliver continuous telemedicine care with zero licensing bottlenecks.

Explore the Ecosystem

```