CONFERENCE / ICCAIS-2026

Modern AI systems have to deal with different types of inputs, such as images, documents, and plain text, within a single system. However, most existing AI solutions are designed as separate pipelines, where users must manually select models and follow different workflows depending on the input type. This paper proposes Omni AI Cloud, a ready-to-use multi- modal AI system that automatically detects the input type and selects suitable AI models using an intelligent routing agent. The system integrates computer vision, natural language processing, document analysis, multilingual translation, and explainability into a single platform. The system activates only the required models and follows a modular design, making it more flexible and easier to deploy than traditional AI systems that are designed to perform a single task.