EduSpeak AI

Inspiration In many classrooms and communities, learners struggle because lessons are too fast, delivered in a language they are not confident in, or lack accessibility tools. I wanted to bridge this gap by building a simple, all-in-one tool that makes spoken content easier to understand, more accurate, and available in multiple languages.

What it does EduSpeak AI takes any spoken input—like a classroom lecture, a meeting, or a voice note—and: Transcribes speech into text (even with accents). Enhances the text by correcting grammar, tense, and clarity. Translates into 20+ languages for multilingual learners. Speaks back the result in a slow, clear, male voice to aid understanding. This helps students, teachers, and under-resourced communities instantly convert spoken information into accessible, accurate, and multilingual learning materials.

How I built it Whisper AI for transcription. Groq LLM for enhancement and translation. Flask for the backend and API endpoints. pyttsx3 + Web Speech API for offline and in-browser text-to-speech. A responsive HTML/CSS UI designed for hackathon-level clarity and accessibility.

Challenges I ran into Ensuring the AI didn’t “over-correct” text and only fixed mistakes. Handling multilingual translation while preserving meaning. Getting consistent slow, male voices across browsers and pyttsx3. Keeping the app lightweight and usable even in low-resource environments.

Accomplishments that I'm proud of Built a working end-to-end pipeline (speech → text → enhanced → translated → speech). Designed an accessible UI that works for both educators and learners Integrated multiple AI models (Whisper + Groq) into one seamless workflow. Created a tool that can genuinely help multilingual and under-resourced communities.

What I learned The importance of limiting AI corrections so that meaning isn’t lost. How to integrate multiple AI services into one smooth web app. The real-world challenges learners face when English is not their first language. Building for accessibility first makes the product more useful for everyone.

What's next for EduSpeak AI Adding real-time live transcription and translation for classrooms. Building a mobile version for low-resource regions.

Built With

• agno
• flask
• groq-llm
• html/css
• javascript
• openai-whisper-api
• python
• pyttsx3
• web-speech-api