1 Head of Department, Department of AI and DS, Kumaraguru College of Technology, Tamil Nadu, India.
2 Department of AI and DS, Kumaraguru College of Technology, Tamil Nadu, India.
International Journal of Science and Research Archive, 2026, 19(01), 414-422
Article DOI: 10.30574/ijsra.2026.19.1.0687
Received on 27 February 2026; revised on 06 April 2026; accepted on 08 April 2026
This research paper gives a voice bot for healthcare domain which utilizes the pipe cat framework for agent orchestration along with Twilio for calling provider / configuration for telephony that encapsulates STT, for conversation management LLM, TTS layer for the conversational AI in real time, state management. At the media layer, stands the Livekit WebRTC for reduced latency and the streaming of audio in bidirectional that achieves about sub 500 ms with an end-to-end response time through SIP trunking.
and Twilio programmable voice API. This bridges the legacy PSTN along with cloud native telephone configuration to optimize the workflow of clinical and the engagement of patients.
Conversational AI; Healthcare Communication; Telephony Systems; Large Language Models (LLM); Speech Recognition; Dialogue Management
Preview Article PDF
Sudharson D, Shruthi P, Afsheen Zaahrah A, Tejaswini SS and Karthick A. Real-time LLM-driven telephony conversational agent for healthcare insurance workflows.International Journal of Science and Research Archive, 2026, 19(01), 414-422. Article DOI: https://doi.org/10.30574/ijsra.2026.19.1.0687.






