Skip to content
Back to Projects

Podcast Creator

Transform PDFs into customizable multi-lingual podcasts with adjustable tone, length, and voice selection using GenAI technologies.

Technical Columbia University

Skills

Generative AI NLP Text-to-Speech Translation

Tools

Python Flask Claude API Google Vertex AI

Podcast Creator transforms PDF documents into customizable, multi-lingual podcasts. Read Less. Learn More. Go Further.

Features

  • PDF-to-Podcast Conversion: Upload any PDF and generate a podcast with customizable parameters
  • Voice Selection: Choose from multiple voice options and tone settings
  • Multi-Language Support: Built-in translation for English, French, German, and Spanish
  • Automated Script Generation: AI-powered script creation with natural dialogue flow
  • Audio Processing: Seamless segment stitching with natural transitions

Technologies

Built with Flask as the web framework, leveraging Anthropic Claude/Sonnet 3.5 for script generation, Google Vertex AI Text-to-Speech API for audio synthesis, and Google Cloud Translation API for multi-language support. Audio processing uses Pydub for stitching podcast segments.

Context

Final project for Columbia University’s E6998 Design for Generative AI course (Fall 2024).