AI as Accessibility: Supporting Neurodivergent Professionals with Customized Systems

This article explores how AI can serve as a crucial accessibility tool for neurodivergent professionals, focusing on a solutions architect with AuDHD (autism and ADHD). The individual developed a personalized AI-powered workflow system using Amazon Quick on desktop, which automates routine tasks such as email triage, task prioritization, and follow-up management. These systems are designed to compensate for executive function gaps, which are common in neurodivergent individuals, and to reduce cognitive load.

The system integrates with Amazon Bedrock for inference and connects to tools like Outlook and Asana through a custom Model Context Protocol server built with AWS Kiro. It classifies and prioritizes emails automatically, flags urgent tasks, and manages task states, allowing the user to start each day with a clear, organized overview. This approach significantly reduces the time spent on administrative tasks, minimizes decision paralysis, and alleviates the mental fatigue associated with traditional organizational tools.

The author emphasizes that AI's role in this context is not about increasing productivity in a typical sense but about making invisible labor visible and automating it away. The system also helps manage emotional burdens such as shame related to missed follow-ups, providing a more sustainable workflow.

This case demonstrates how AI can be tailored to meet specific cognitive needs, offering a model for accessible technology that can benefit many neurodivergent individuals. While the system is personalized and not a commercial product, it highlights the potential for AI to improve quality of life and work efficiency for users with similar challenges.

The development of such systems could influence future AI tools aimed at accessibility, encouraging more inclusive design practices and personalized solutions that address diverse cognitive profiles. It also underscores the importance of user-centric AI development in creating practical, supportive applications.