Patent Position & Technical Mapping

Amazon Web Services operates an extensive visual, multimodal and generative-AI architecture across Amazon Bedrock, Amazon Nova, Amazon Rekognition and associated AWS data and agent services.

The publicly documented AWS architecture includes capabilities relevant to several central functions of the patent position, including:

• image and video analysis;
• object and scene identification;
• visual question answering;
• multimodal understanding;
• visual and textual input processing;
• multimodal information retrieval;
• contextual response generation;
• connection of visual understanding to external information and tools;
• AI-generated assistance;

AWS has also recently published an “agentic vision” architecture expressly combining computer vision, agents and Model Context Protocol infrastructure into a coordinated visual-information pipeline.

The technical mapping below evaluates these functions against the patent-position architecture using primary AWS documentation.