About Selora
Selora is a tool for practicing pronunciation: Spanish if you speak English, English if you speak Spanish. You listen to a reference, see the syllable-by-syllable breakdown, and practice with your own voice.
About the creator
Selora is a project by Sebastián Vergara Gac, a Product Designer based in Porto. It's a personal experiment exploring how to make pronunciation practice feel clear and friendly — built gradually, honestly, and in the open.
How it's built
Under the hood, Selora runs on voice AI: speech synthesis generates the reference audio, and pronunciation-recognition models analyze your voice syllable by syllable. The goal is to turn that into useful, instant feedback without getting in the way of the experience.
Each word gets its own server-rendered page with its phonetic transcription and syllable breakdown. When you record, your voice is captured and processed in the browser, compared against a native reference, and turned into a per-syllable score shown as a three-color traffic light. The priority is keeping the technology out of the way, so practicing feels simple.
Sources
Word definitions come from Wiktionary, trimmed to fit here and translated where needed. They're used and shared under the CC BY-SA 4.0 license. Reference audio and voice analysis are generated by Azure Speech.
Feedback
Got an idea, or found something broken? Tell me in thirty seconds — no sign-up, and the email is optional.