{
  "id": 1024060,
  "title": "Building Arogya Seva: How I Built an Ultra-Low Latency Telehealth Voice AI for Bharat in 10 Days",
  "url": "https://urgent.news/2026/08/15/building-arogya-seva-how-i-built-an-ultra-low-latency-telehealth",
  "topic": "health",
  "section": "Health & Medicine",
  "published": "2026-08-15T11:20:47.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/viral1998/building-arogya-seva-how-i-built-an-ultra-low-latency-telehealth-voice-ai-for-bharat-in-10-days-1cmg"
  },
  "original_language": "en",
  "account": "Across rural and semi-urban India, millions of citizens delay seeking medical advice due to high patient-to-doctor ratios, geographical distances, and language barriers. Arogya Seva, launched as part of the #VoiceForBharat challenge, aims to bridge this gap with an empathetic, multilingual, real-time voice assistant.\n\nThe system's architecture focuses on ultra-low latency to deliver a natural conversational experience. Speech-to-text transcription is performed in real-time using Deepgram Nova-3, while Google Gemini 2.0 Flash processes intent and applies clinical guardrails. Text-to-speech synthesis is handled by Murf Falcon, providing warm, empathetic, and human-like voice output.\n\nOver a 10-day period, Arogya Seva was built with several key features. First, it offers sub-second voice synthesis powered by Murf Falcon, ensuring fast response times. Second, strict guardrails and native script enforcement ensure safety and proper acoustic synthesis, with Hindi responses strictly in Devanagari script. Third, dynamic frontend state and audio visualizers display real-time agent states using Next.js and LiveKit Agents UI.\n\nPrivacy is paramount, with explicit consent required for storing call data and the option to opt-out at any time. Real-domain health tools and tool chaining automate health triage and clinic appointment management. When needed, the system can escalate calls to human specialists, storing sanitized call summaries for reference.\n\nThe project faced challenges such as TTS latency spikes and accidental code-mixed script bleed. These were addressed by integrating Murf Falcon with streaming tokenization and implementing a system prompt guardrail, ensuring smooth and safe operation.",
  "summary": "Introduction: The Problem & The Mission Across rural and semi-urban India, access to timely healthcare advice is hindered by high patient-to-doctor ratios, geographical distances, and language barriers. Millions of citizens delay seeking medical advice simply because scheduling a clinic visit or understanding preliminary symptom urgency is daunting. Arogya Seva was created to bridge this gap as…",
  "key_points": [
    "Arogya Seva aims to provide ultra-low latency telehealth voice AI for rural India.",
    "System uses Deepgram Nova-3, Gemini 2.0 Flash, and Murf Falcon for real-time processing.",
    "Privacy ensured with explicit consent, script enforcement, and call data opt-out."
  ],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}