AI Visual Reasoning Assistant

AI Visual Reasoning Assistant

An AI-powered accessibility assistant that combines voice, computer vision and natural conversation to help users better understand and interact with the world around them.

An AI-powered accessibility assistant that combines voice, computer vision and natural conversation to help users better understand and interact with the world around them.

Category

AI Assistant / Computer Vision / Accessibility Tech

AI Assistant / Computer Vision / Accessibility Tech

Services

Kotlin, Jetpack Compose, Node.js, TypeScript, Google Cloud(Gemini, Vision, Voice), SQLite, REST APIs

Kotlin, Jetpack Compose, Node.js, TypeScript, Google Cloud(Gemini, Vision, Voice), SQLite, REST APIs

Year

2025/26

2025/26

  • AI-powered accessibility assistant designed to help users better understand and interact with the world around them.

  • Combines natural conversation, computer vision and voice interaction into a single experience.

  • Analyses the user's surroundings and answers questions in real time using AI.

  • Maintains conversational context to provide more natural and personalised interactions.

  • Supports speech recognition, AI reasoning and natural voice responses.

  • Built with a modular client-server architecture for scalability and future expansion.

  • Designed with accessibility, simplicity and hands-free use as core principles.

  • Development evolved through multiple prototypes, from smart glasses to a Raspberry Pi proof of concept before arriving at the current Android implementation.

  • Created as the foundation for future wearable accessibility devices, including smart glasses.

  • Focused on delivering fast, reliable and intuitive everyday assistance.

Prototype 3 — Android assistant app

Prototype 2

Prototype 2 — Raspberry Pi assistant

Prototype 1

Prototype 1 — Smart glasses

Create a free website with Framer, the website builder loved by startups, designers and agencies.