This is a full-stack web application built to demonstrate how modern AI can be used to automatically analyze and pull useful information from a simple phone call recording.
The app takes a sample audio file and performs three key AI tasks:
- Transcription: It uses OpenAI's Whisper to get a full, accurate text transcript of the conversation.
- Sentiment Analysis: It analyzes the text to figure out if the overall tone of the call was positive or negative.
- Summarization: It creates a quick summary so you can understand what the call was about without reading the whole thing.
This project combines a Python backend for AI processing with a modern JavaScript frontend.
Backend:
- Language: Python
- API Framework: FastAPI
- Transcription Model: OpenAI Whisper
- Analysis Models: Hugging Face Transformers
Frontend:
- Framework: Next.js
- Language: TypeScript
- Styling: Tailwind CSS