As a developer, I want a centralized Tailwind CSS theme configuration with the project's color palette (#0D1B2A background, #1B263B surface, #E0E1DD text, #FF6F61 accent, #415A77 muted) so that all components share a consistent cinematic visual identity.
As a learner, I want my text or voice-transcribed questions sent to a FastAPI endpoint that returns AI-generated response text so that the tutor can answer my questions intelligently.
As a guest user, I want to see a timezone-consistent Daily Power Fact on the Landing page that rotates daily so that each visit offers a fresh and relevant learning insight.
As a guest user, I want to see the Landing page rendered faithfully from the existing JSX design (v2) including hero section, navigation, Daily Power Fact teaser, and call-to-action so that I understand the product and am compelled to try the Demo.
As a learner, I want to see the Demo page rendered faithfully from the existing JSX design (v2) including the chat interface, avatar viewport, subtitle panel, waveform area, and voice input controls so that I can immediately begin interacting with the AI tutor.
As a learner, I want the AI tutor's text responses converted to realistic speech audio via ElevenLabs TTS so that I can hear the tutor speak in a professional, natural-sounding voice.
As a guest user, I want to see an animated background particle and grid system on the Landing and Demo pages so that the interface feels futuristic and cinematic without impacting performance.
As a guest user, I want cinematic scroll-triggered animations on the Landing page sections so that the experience feels immersive and professionally produced as I explore the page.
As a learner, I want to see a realistic 3D AI tutor avatar rendered via Three.js with a loaded GLB model supporting idle and talking animation states so that the interaction feels like speaking to a real news-anchor style instructor.
As a learner, I want to speak my question directly into the browser using the Web Speech API so that I can interact with the AI tutor hands-free without typing.
As a guest user, I want the Landing page to fetch and display the current Daily Power Fact from the backend API so that the content is dynamic and timezone-aware.
As a learner, I want the 3D avatar's mouth to synchronize with AI-generated audio using viseme sequences and morph-target blending so that the tutor's speech appears natural and realistic.
As a learner, I want to see a real-time animated audio waveform during voice input and AI speech playback so that I have clear visual feedback of the active audio state.
As a learner, I want my typed or spoken question to flow through voice input, the chat API, ElevenLabs TTS, avatar lip-sync, and subtitle display in a seamless end-to-end interaction so that the Demo page functions as a complete AI tutor session.
As a mobile learner, I want the Demo and Landing pages to implement lazy-loading for heavy assets and a reduced-motion fallback for animations so that the experience is smooth and accessible on lower-powered devices.

Voice se pucho, AI se samjho
From voice to knowledge in four simple steps — experience AI-powered learning like never before.
Use your voice to ask anything — the Web Speech API captures your words naturally in Hindi or English.
Our AI tutor processes your question and generates a clear, structured response in text and voice.
A realistic 3D avatar speaks the answer back to you with synchronized lip movements and natural gestures.
Absorb knowledge through an immersive, cinematic experience that makes complex concepts easy to understand.
Ask questions naturally using your voice. Powered by the Web Speech API, gentle-avatar understands both Hindi and English for a seamless bilingual experience.
Watch the 3D avatar speak with perfectly synchronized mouth movements using viseme and morph-target cues for a lifelike tutoring session.
Get intelligent, context-aware answers generated by advanced AI. Receive both text and voice responses tailored to your learning needs.
Enjoy a premium experience on any device. Lazy-loaded 3D assets and reduced-motion fallbacks keep performance smooth on mobile.
“The human brain processes speech at roughly 150 words per minute, but can comprehend up to 400 words per minute when listening — making voice-based AI tutoring one of the fastest ways to absorb knowledge.”
July 11, 2026 · UTC
Real experiences from students and professionals using gentle-avatar
Experience the future of interactive learning. Ask questions by voice, get instant AI-powered answers, and engage with a realistic avatar tutor — all in one place.
No sign-up required • Free to explore • Works on all devices
No comments yet. Be the first!