video1.0<iframe src="https://www.loom.com/embed/0c994429980741b0ae3e603ec8562043" frameborder="0" width="1920" height="1440" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>14401920Loomhttps://www.loom.com14401920https://cdn.loom.com/sessions/thumbnails/0c994429980741b0ae3e603ec8562043-d0c652e32c92823c.gif292.397In this video, I share my experience building an automated voice bot for healthcare using Telnyx, OpsGenie, OpenAI Realtime, and FastAPI. I chose Telnyx for its privacy features and higher performance compared to Twilio, which is crucial for handling sensitive patient information. I discuss the improvements made by switching to OpenAI Realtime for more natural conversations and highlight significant bugs found during testing, such as issues with patient record retrieval and language barriers. I suggest that agents should prioritize patient needs over identification and refer urgent cases to the ER immediately. If I had more time, I would enhance the system by adding more edge cases and refining patient profile scenarios.