Iman-301/ai-farmer-advisor

Voice-driven Amharic farming assistant that connects farmers with reliable agricultural knowledge. Uses Whisper for speech-to-text, RAG for contextual answers, and gTTS for natural voice replies. Includes a Streamlit dashboard and containerized microservices for easy deployment and scaling. # Voice-To-Voice Farmer Advisor System LINK TO THE DEMO This project implements a Voice-to-Voice Farmer Advisor System that provides agricultural information to farmers in Amharic via a voice-based interface. The system uses speech-to-text (STT) to understand queries, a logic service with RAG (Retrieval-Augmented Generation) to find relevant information, and text-to-speech (TTS) to deliver responses. ## 🚀 Features - **Amharic Speech-to-Text**: Uses OpenAI's Whisper model for accurate transcription. - **RAG-Based Logic Service**: Retrieves information from a knowledge base using vector similarity search. - **Amharic Text-to-Speech**: Generates natural-sounding Amharic speech. - **Admin Dashboard**: A web interface to view escalation queues and manage the system. - **Dockerized Deployment**: All services are containerized for easy deployment and management. ## 🛠️ Architecture The system consists of four main services: 1. **Telephony Service**: Handles incoming calls, audio capture, and playback. 2. **STT Service**: Transcribes audio to text. 3. **Logic Service**: Processes text queries, performs RAG, and generates responses. 4. **TTS Service**: Converts text to speech. 5. **Admin Dashboard**: A web interface for system monitoring and management. ## 📂 Project Structure ``` Voice-To-Voice-Farmer-Advisor/ ├── telephony_service/ # SIP/VoIP client and audio processing ├── stt_service/ # Speech-to-text service ├── logic_service/ # RAG pipeline and business logic ├── tts_service/ # Text-to-speech service ├── admin_dashboard/ # Streamlit-based admin interface ├── data/ # Data files (knowledge base, embeddings) ├── .env # Environment variables └── docker-compose.yml # Docker orchestration ``` ## ⚙️ Prerequisites - Docker - Docker Compose ## 🚀 Quick Start 1. **Clone the repository** (if not already done). 2. **Configure Environment Variables**: Copy the example environmen …