A spiritually guided chatbot offering divine wisdom inspired by the Bhagavad Gita.
Sarathi is a spiritual chatbot designed to simulate conversations with Shree Krishna, offering responses rooted in the teachings of the Bhagavad Gita. Built with modern web technologies and powered by a Language Model (LLM), the chatbot allows users to ask personal or philosophical questions and receive compassionate, scripture-inspired answers.
- Divine responses based on Bhagavad Gita verses.
- Voice response (text-to-speech) in both English and Hindi.
- Language toggle support (English ๐ฌ๐ง / Hindi ๐ฎ๐ณ).
- Krishna-style emotional welcome and response tones.
- Typing animation mimicking divine response generation.
- Scrollable, modern chat UI with speech synthesis.
- Intelligent backend powered by LLM + Pinecone + Gemini.
Frontend:
- React with TypeScript
- Framer Motion for animations
- Tailwind CSS
- React Markdown
- Lucide React icons
- Web Speech API (Text-to-Speech) Backend:
- Flask (Python)
- Gemini LLM API
- Pinecone Vector DB
- Render for deployment
git clone https://github.com/Vrajesh-Sharma/sarathi.git
cd sarathiInstall frontend dependencies
cd frontend
npm install
npm run devRun backend server
cd backend
pip install -r requirements.txt
python app.py๐ Note: Make sure your .env contains valid keys for Pinecone and Gemini API.
This project is hosted using:
- Frontend: Vercel - Sarathi
- Backend: Render - To deploy your own version, update environment variables and deploy with your preferred platforms.
- Open the web app.
- You'll be greeted by Krishna with a warm spiritual message.
- Ask any question about:
- Life struggles
- Work and business dilemmas
- Family issues
- Spiritual confusions
- Krishna responds based on Bhagavad Gita teachings.
Example Questions:
- โIโm feeling lost in my career. What should I do?โ
- โHow do I handle separation in a marriage?โ
- โWhat does Krishna say about material desires?โ
โYou have the right to perform your actions, but not the fruits of them.โ
โ Bhagavad Gita 2.47
Made with โค๏ธ by Vrajesh Sharma