| dc.description.abstract | Managing doctor appointments in healthcare remains challenging due to long wait times,
manual triage errors, and scheduling conflicts. This thesis presents a Smart Doctor
Appointment Agent leveraging workflow automation and conversational artificial
intelligence to streamline hospital scheduling.
Built on the n8n automation engine, the system integrates the OpenAI Chat Model
(gpt-5-mini), Google Sheets, and Google Calendar. Operating via a chat interface, the AI
agent parses unstructured patient inputs and decodes phonetic symptom descriptions
across Bangla, English, and Banglish (e.g., "buke betha"). Using deterministic function
calling, the agent searches a Google Sheets database to match symptoms with specialized
doctors and chamber schedules.
To prevent double-booking, the workflow performs real-time calendar availability checks
locked strictly to the Asia/Dhaka (+06:00) timezone. After collecting patient metadata
and receiving final user confirmation, the agent automatically creates a live appointment
entry in Google Calendar.
By orchestrating AI reasoning with backend APIs, this low-code architecture operates
24/7, reduces administrative overhead, and eliminates schedule drift, delivering an
efficient, friction-free booking experience for modern healthcare administration. | en_US |