Podify

Podify: Turn any article into a custom podcast. Upload a URL, generate a personalized podcast experience, and get an instant summary of the key takeaways from it.

Comment

Podify — Turning Links into Personalized Podcasts

Inspiration

Ever wanted to squeeze in a bit more studying while walking to an exam, going for a drive, or heading out on a long walk? I constantly found myself bookmarking articles and videos I meant to get back to, but never did.

I learn best through repetition and passive listening, especially during commutes or workouts. I wanted a way to turn content I was already saving into something I could actually consume during those in‑between moments.

That motivation led to Podify, a tool that lets you paste in a link and instantly turn it into a podcast‑style audio experience. Instead of pushing learning back, Podify helps you learn while you’re already moving.

What I Learned

Building Podify taught me lessons across product, engineering, and AI system design:

  • Product framing matters as much as the model: small UX decisions (like how you ask for user intent) drastically change perceived value.
  • Audio-first experiences require different writing: scripts must sound natural when spoken, not just read.
  • Designing for extensibility early pays off: treating user input as structured intent (not raw prompts) made the system easier to evolve.

How I Built It

Podify is built as a full‑stack web application focused on fast iteration and clean separation of concerns.

Tech stack:

  • Backend: FastAPI (Python)
  • Frontend: React + TypeScript
  • Styling: Tailwind CSS
  • AI: Google Gemini (script generation)
  • Audio: ElevenLabs (text‑to‑speech)

Architecture Overview

  1. The user submits a link (article or YouTube video) through the React frontend.
  2. The FastAPI backend handles content ingestion and parsing.
  3. Cleaned text is passed to Google Gemini to generate a podcast‑style script.
  4. The generated script is sent to ElevenLabs to produce natural‑sounding audio.
  5. The frontend plays the resulting podcast immediately.

A key feature is the optional user context input, which allows listeners to specify what they want out of the content (e.g., comparison, actionable advice, or learning fundamentals).

What’s Next

The future of Podify focuses on making the listening experience even more flexible, personal, and expressive.

One direction is expanding voice and style options using ElevenLabs. Instead of a single narrator, users will be able to choose from multiple voices and tones to simulate different podcast styles. Such as educational, conversational, or storytelling, so the audio better matches how they prefer to learn.

Furthermore, beyond articles and links, Podify will support YouTube videos and PDFs, allowing users to turn lectures, papers, and long‑form documents into podcasts they can listen to on the go.

Built With

Share this project:

Updates