Tavanito

The problem

Where the barrier is.

Mainstream digital interfaces are dense, fast, and visually driven. People who rely on screen readers or have cognitive or motor disabilities spend enormous effort on tasks that should be trivial.

What it does

How it helps.

A conversational layer that sits between the person and their digital tasks, translating intent into action with explicit, reversible steps and full screen-reader support throughout.

The impact

What it changes.

Beta users report completing tasks — like booking appointments or replying to email — in a fraction of the time, with far less frustration and without asking another person for help.

In depth

How it works.

Tavanito is a conversational assistant designed specifically around accessibility, not retrofitted for it. People describe what they want in natural language — fill in this form, summarize this page, read my messages, help me reply — and Tavanito carries it out while narrating each step. It is built to be screen-reader native, keyboard-first, and patient. Where most AI assistants optimize for speed, Tavanito optimizes for clarity, confirmation, and control, because for many users a wrong action is far costlier than a slow one.

Built with

• Next.js
• TypeScript
• LLM Orchestration
• ARIA
• Web Speech API