Hey Product Hunt! 👋 I’m Aleksei, founder and mobile engineer at ReactLab, building alongside our compact team of 4 (Denis on mobile, Buin on backend, and Kseniia on design) out of Tallinn, Estonia 🇪🇪. Today, we’re excited to launch Edyssey on iOS and Android! The Problem We Wanted to Solve Most language apps force everyone down the exact same rigid, generic lesson tree. You waste time tapping words you already know, get hit with red penalty screens and lost "hearts" without contextual grammar explanations, and hear robotic text-to-speech. Meanwhile, first-gen AI chat wrappers suffer from "tutor amnesia"—every conversation starts from zero, completely forgetting what you struggled with yesterday. How Our Approach Evolved Inspired by the Michel Thomas method and real 1-on-1 human tutoring, we built our CORE_v4 engine around a simple realization: fluency comes from targeted, stateful correction over time. Shaped by feedback from our first ~1,000 early learners, we evolved Edyssey around Token-Level Adaptive Memory and low-latency Gemini Live voice conversations: 🧠 100% Adaptive Token-Level Memory: Every attempt you make is analyzed in real time and categorized as Struggling, Learning, or Mastered. When you miss a grammar nuance or hesitate on a word, Edyssey saves it to your personal memory bank and automatically weaves it into upcoming speaking, reading, listening, and writing challenges until it sticks. 🎙️ Low-Latency Voice Dialogue: Practice fluid, native-speed conversations across 35+ languages (A1–C1) with instant pronunciation scoring and gentle, plain-English grammar breakdowns—without stage fright. 📚 5-Way Dynamic Sessions: Short 10-minute sessions that combine AI voice speaking, native-speed listening, smart sentence construction with particle/letter error highlighting, and level-matched reading stories. No two sessions are ever identical. 🔒 Privacy-First Voice Processing: Your voice audio is processed in real time solely to evaluate pronunciation and is never stored on our servers after the session ends. Edyssey is live right now on the App Store and Google Play, and you can jump straight into your first session out of the box with zero friction. We’d love your feedback: Which of the 35+ languages are you testing today, and how does the voice cadence feel? What specific speaking scenario or grammar breakdown would make your daily 10-minute session even better? Denis, Buin, Kseniia, and I will be in the comments all day answering your technical and product questions! 🚀

Edyssey

Adaptive AI language tutor that remembers your weak spots

Adaptive AI language tutor that remembers your weak spots