HerEcho

HerEcho is an AI-powered voice app advancing SDG 5 by providing personalized audio education to illiterate women in underrepresented regions for accessible, empowering learning.

Comment

🪶 HerEcho: Where Every Voice Learns

💡 Inspiration

770 million adults* around the world can’t read — and two-thirds are women.

In low-income countries, female literacy rates average ~54%, compared to ~70% for men. Traditional education assumes you can read, type, or own a screen — but millions are left behind simply because learning still depends on literacy.

We asked ourselves: What if education didn’t require reading or screens at all?

That question became HerEcho, an AI-powered voice learning app that teaches through simple, human conversation.

🎧What it does

HerEcho = Learning by Voice

  1. Speak a question.
  2. The AI listens and understands.
  3. The AI replies with an audio lesson — in your own language.

Key Features 🗣️ Audio-only learning (no reading, no typing) 🌍 Multilingual support for local and low-resource languages 🔊 Spoken resources (NGO info, events, training) 🔁 Natural follow-up conversations — all through voice

HerEcho: Where Every Voice Learns Education without reading. Inclusion without screens. Empowerment through voice.

🔧 How we built it

Frontend (React)

Built a WhatsApp-style interface with: 🎤 Mic button for recording 🔴 Recording animation 🔊 Audio playback bubbles Integrated “audio resource cards” that summarize resources verbally

Backend (Node.js + Express)

Audio pipeline: Whisper → GPT → Text-to-Speech Endpoint /audio-reply returns: AI lesson audio Spoken resource audio Preloaded audio ensures real-time demo performance

Tech Stack

React Node.js Express OpenAI GPT Text-to-Speech APIs

🚧Challenges we ran into

Whisper + TTS latency for real-time conversations Handling multiple languages under limited bandwidth Designing a screenless flow (no text fallback!)

🏆Accomplishments that we're proud of

Built a working React + Node.js MVP in under 24 hours Integrated Whisper, GPT, and Text-to-Speech Designed a fully audio-first UI/UX Created pre-recorded voice lessons and resource audios

🚀 What’s Next

🌐 Expand multi-language and offline support 📞 Build HerEcho Dial-In — phone-based version 🎙️ Add personalized voices and adaptive tone 🤝 Partner with NGOs for pilot deployment 🧠 Create HerEcho for Educators — voice-based lesson creation

Links

Github Presentation

Built With

Share this project:

Updates