Jed Lee developed a real-time vision feature for the alibaba/MNN repository, enabling users to capture and transmit images during interactive voice chat sessions. Leveraging Android development skills and the CameraX library, Jed integrated live camera functionality directly into the chat interface, allowing seamless image sharing alongside voice communication. The implementation included multimedia processing and custom UI/UX design to provide intuitive camera controls within the chat workflow. This work established an end-to-end media capture path for the MNNChat feature, addressing the need for richer collaboration and laying the technical foundation for future video-enabled interactions. The focus remained on robust feature delivery.