SpeakEasy - AI Voice Agent

SpeakEasy is a real-time, multilingual AI voice assistant designed to make voice-based interaction simple, natural, and accessible. The system allows users to communicate through a browser using their microphone and receive spoken responses from an AI voice agent. SpeakEasy supports English, Hindi, Marathi, and conversational Hinglish interactions. It uses AssemblyAI's Voice Agent API for real-time speech processing, conversation management, and intelligent tool calling. The key feature is its ability to understand an action request and select the appropriate tool. For example, when a user says "Turn on the fan" or "Turn off the light," the AI identifies the required action and calls the corresponding Python function. The current implementation includes control_fan and control_light tools that maintain the device states and return the execution result to the voice agent. The project includes a Python backend, browser-based frontend, real-time audio processing, temporary token authentication, and a deployed web interface. GitHub is used for source-code management and Render is used for cloud deployment. The architecture is designed to be extensible. In the future, the Python tools can be connected to real IoT hardware such as ESP32-based appliances, allowing SpeakEasy to evolve from a software demonstration into a practical voice-controlled smart-device system.