TRANSCRIPTO

Index page
Register
Login
Dashboard page

Inspiration

We were inspired by the need for a simple yet powerful tool that transforms voice into usable insights. From busy professionals to people with auditory disabilities, many benefit from automated transcription and summarization. We wanted to build a solution that is both accessible and intelligent.

What it does

Transcripto converts audio into readable text and summarizes the content using AI. It allows users to upload audio files or record live speech and then receive both a full transcription and a concise summary—making it ideal for note-taking, podcasting, research, and accessibility.

How we built it

We used the Flask web framework for the backend, integrating the Google Speech Recognition API for transcription and Hugging Face transformers (BART) for summarization.
The frontend was built with HTML, **CSS, and **JavaScript.
We also used PyDub and FFmpeg for audio processing and format conversion.
All user interactions take place through a clean, modern UI with real-time feedback.

Challenges we ran into

Audio quality varied greatly between files, affecting transcription accuracy.
Summarization models required minimum input length and internet access for downloading.
Integrating live recording in-browser across devices was complex.
FFmpeg compatibility issues across operating systems posed deployment difficulties.

Accomplishments that we're proud of

Seamless integration of transcription and summarization pipelines.
User-friendly UI supporting both live and uploaded audio.
Successfully handled multiple audio formats including WAV, MP3, and FLAC.
Fully functional prototype with clear and concise output.

What we learned

Real-time audio processing requires careful consideration of latency and browser compatibility.
Hugging Face’s models are powerful but need preprocessing and threshold handling for best results.
Even a simple authentication system improves usability and structure.
Small touches in UI design greatly improve the user experience.