AI Clinic Assistant

🏥 On 15 July 2025, this project was formally proposed as a pilot to the NHS Innovation Hub.

📄 View full proposal email (PDF) or 📷 see email screenshot 

 Introduction

The AI Clinic Assistant is a fully operational, voice-based application designed to automate appointment handling in healthcare clinics. It conducts real-time, natural conversations with patients over standard landlines — without requiring cloud telephony (e.g., Twilio) or technical staff.

Originally piloted in a private clinic in London run by an NHS-affiliated doctor, the assistant behaves like a human receptionist, allowing patients to book, reschedule, or cancel appointments by calling the clinic’s regular phone number.

Problem It Solves

Many small and mid-sized clinics lose time and efficiency by handling appointment-related calls manually. Staff are overwhelmed with repetitive tasks, especially during busy hours — and there's no scalable, plug-and-play solution for 24/7 phone-based appointment handling.

AI Clinic Assistant fills this gap by offering an easy-to-deploy, affordable, and reliable voice automation solution.

 Key Features

•  Natural voice interaction: Understands and responds to patients like a real receptionist
  
  

•  No cloud telephony required: Works offline on a local PC, using a standard landline
  
  

•  Zero technical setup: Clinics can use it without IT teams or programming knowledge
  
  

•  Scalable design: Adaptable for GP clinics, dental practices, therapy centres, etc.
  
  

•  24/7 coverage: Can handle calls outside working hours or during lunch breaks
  
  

 Project Status

•  Successfully tested in a real clinical setting with NHS-affiliated medical staff
  
  

• Refined through feedback from clinic managers, receptionists, and doctors
  
  

• Multiple versions developed with feature improvements and iteration tracking
  
  

 Screenshots & Demo Video

▶️ Watch Full Demo

 Screenshots (suggested visuals to upload):

• Incoming call & speech recognition
  
  

• Verbal booking confirmation
  
  

• GPT-powered conversation interface
  
  

 GitHub Repositories

• clinic-assistant-v1: (March 2025) Basic logic for voice-to-text transcription and Google Calendar syncing.
  
  

• clinic-assistant-v2: (April 2025) Enhanced error handling, improved name/date/time extraction, and more stable appointment creation.
  
  

• clinic-assistant-v7: (July 2025) Final polished version used in the public demo — supports full voice interaction with GPT, natural flow tracking, and appointment confirmations.

• clinic-assistant-v8:(August 2025) Plug-and-play USB modem integration for real-time landline call detection and automatic voice response — ready for deployment in live clinic environments.

• clinic-assistant-v9: (October 2025): Mobile-first Android release with real-time auto-answering, natural GPT-powered conversations, and direct Google Calendar integration — removes PC and hardware dependencies, delivering true plug-and-play capability for every clinic.
  
  

 Planned Features (Next Versions)

The following enhancements are planned to evolve the system into a fully scalable, user-friendly, and cloud-ready product:

• Plug-and-Play Phone Line Integration: Automatic detection and response to incoming calls when a standard landline is physically connected to the PC. The assistant starts in real time upon call arrival.
  
  

• Interactive Web Dashboard: A web-based interface that displays upcoming appointments in a calendar-style view, categorized by treatment type. Hovering over a slot reveals additional patient details.
  
  

• Customizable Voice Interaction Settings: Users will be able to personalize greeting and closing messages via a simple web-based configuration menu.
  
  

• Calendar Data Import/Export: Ability to import Google Calendar data into the dashboard (pre-exported .ics or .csv format) and export the same back from the interface for backup or migration purposes.
  
  

• Centralized Backup Directory: All imported/exported data will be securely stored in a dedicated backup folder for easy restoration and data safety.
  
  

• Google Login & Calendar Sync: Automatic Google sign-in from the dashboard to sync calendars without manual authentication steps.
  
  

• Call Analytics Module: An analytics section to display categorized call statistics, e.g., appointment-related, unrelated, or requiring human follow-up, with timestamped logs.
  
  

• Secure API Key Handling: Integration of GPT or calendar APIs through encrypted storage, ensuring that API keys are never exposed to end users.
  
  

• User Account Management: Dashboard login tied to user-specific accounts, supporting personalized configurations and usage limits.
  
  

• Usage-Based Billing: The system will calculate and reflect billing amounts based on API usage thresholds, with automated balance adjustments.
  
  

• Cloud-Native Architecture: The entire system will be designed for deployment on cloud infrastructure (e.g., Firebase, AWS, or Google Cloud), ensuring high availability, scalability, and security.

These planned features aim to make the assistant not just functional, but also professionally deployable at scale across sectors such as healthcare, legal services, and other client-based industries.

Last updated: 2025-07-03

Timeline

This page presents the development timeline of my AI Clinic Assistant and related innovation milestones.

From the initial idea sparked by real-world problems in a clinic, to working prototypes, public showcases, and future plans — each phase reflects a hands-on, purpose-driven approach to building scalable, human-friendly AI systems.

By laying out these key stages, this timeline demonstrates that the project was not built for a visa application — but rather, the visa supports a project that was already in motion.