Signal LLM Chatbot

For this project, I developed an advanced Signal-based personal assistant chatbot using Signal-cli, designed to operate entirely within Signal Private Messenger while offering a wide range of intelligent, multimodal capabilities.

The assistant integrates several OpenAI APIs, enabling natural, contextual conversation through GPT models, text-to-image generation with DALL·E, and speech-to-text transcription powered by Whisper. These capabilities allow the bot to respond conversationally, generate images on demand, and convert voice messages into text.

To extend functionality beyond conversation, I integrated a suite of REST APIs that provide real-time access to weather data, satellite imagery, flight tracking, and IoT device controls—allowing the assistant to fetch information or directly interact with smart-home systems.

For computer-vision capabilities, the bot uses OpenCV alongside a YOLOv8 Convolutional Neural Network(CNN) for object detection model, enabling it to perform object recognition and tracking. Detected objects and annotated results are delivered directly to the user through Signal, turning the chatbot into a versatile vision-enabled assistant.

This combination of multimodal AI, computer vision, and device integration results in a highly capable, privacy-preserving personal assistant fully contained within the Signal ecosystem.

Images