What I Built

Sanjeevani is a multilingual AI-powered virtual doctor that accepts voice, text, or image inputs and responds with realistic, human-like diagnosis and remedy using Groq's LLM and Murf AI.

It addresses the problem of language barriers and accessibility in digital healthcare. Whether a patient speaks Hindi, French, Spanish, or Chinese — Sanjeevani listens, understands, and speaks back like a real doctor.

---

Demo

🎥 Watch Sanjeevani in action:

---

🔗 Code Repository:

https://github.com/this-is-rachit/Sanjeevani

---

🌐 Try it live:

https://sanjeevani-6vck.onrender.com/

---

🔊 How I Used Murf API

Murf AI powers the voice and translation layer in Sanjeevani:

• ✅ Text-to-Speech: Converts Groq-generated medical advice into lifelike speech using Murf’s voice models.
• 🌐 Multilingual Translation: Automatically translates diagnosis into the selected language before speech synthesis.
• 🎙️ Voice Mapping: Used Murf's voice IDs to customize the sound per language (e.g., Hindi, Japanese, German).

This brings a human warmth to AI conversations — vital for a healthcare app.

---

💡 Use Case & Impact

Real-World Applications

• 🏥 Rural/Remote Healthcare: For patients who can’t read or write, Sanjeevani offers voice-based, language-native assistance.
• 🌍 Global Accessibility: With 16+ language support, it’s usable from India to Italy.
• 🖼️ Image Support: Users can upload a rash or wound image for visual diagnosis via LLM.

Impact

Sanjeevani enhances digital healthcare accessibility, especially for non-English speaking and underserved populations. It’s a step toward inclusive AI in medicine.

---

🧪 Tech Stack

• Murf AI – Text-to-speech and multilingual translation
• Groq LLaMA 4 – Medical advice generation
• Groq Whisper – Voice-to-text transcription
• Gradio – Web-based interface
• Python, langdetect, PyDub, SpeechRecognition

---

🧠 Built solo by @rachit_bansal