Nativo.
Practice English with your own texts, browser-generated audio, and writing feedback.
Overview
Nativo brings reading, listening, vocabulary, and writing practice into one application. You choose the material: a passage from a book, an article, or any text you want to study. A speech model downloaded into the browser generates the audio locally.
How it works
Create a lesson by pasting its content and adding a title, description, and optional source information. Generate its audio to listen to the lesson or individual sentences while reading in sentence or page view. The generated audio is saved for later playback.
Words are highlighted according to how well you know them. Select a word or phrase to see its contextual meaning and pronunciation, hear it aloud, or add a note. Mark words as ignored or assign a familiarity level; known words stop being highlighted. An optional English-level setting initializes familiar vocabulary from earlier levels so common words do not clutter the reading view. You can also reveal a Spanish translation.
Expression detection is optional. Run it on a lesson to identify and save phrasal verbs and idioms for study. For writing practice, create a document and use Check to review grammar and coherence feedback, accepting or rejecting individual corrections.
Technical decisions
Paste content directly
I chose copy and paste rather than automatic article scraping. It gives the user control over the exact passage and avoids unreliable extraction and formatting across different websites.
Detect expressions, then validate their context
I split lesson text into chunks and extract idioms and phrasal verbs in separate calls. A second stage validates each candidate against the surrounding text to reduce false positives. Accepted and rejected results are stored for review, and I use reference examples and a comparison script to evaluate changes.