AudiowaveAI

Inspiration

It started with me buying a book -- Make by Pieter Levels -- which I wanted to convert to an audiobook. Existing tools were either too expensive or sounded horrible. One evening I tried to create my own and posted about it on Twitter/X. People really liked it and asked me to make it into a real product.

A few months later I did just that.

What it does

Converts text to high-quality audio using the latest AI models and techniques. It works on books, articles, blog, posts, pdfs, epubs, URLs or any text.

How I built it

Hacked on it for a few days to get the core, posted about it and got feedback. Then started using it myself and kept improving for a few weeks.

Challenges I ran into

Creating audio with AI is expensive and took a while to find the right models to use. It also only works for shorter text so we need to chunk the content up and stitch it back together at the end.

Getting traffic and letting people know is the biggest challenge.

Accomplishments that I'm proud of

Launched on Product Hunt and Hacker News and it went viral and started to be used by a lot of people.

What I learned

How to move fast and iterate, a lot of learning about FFMPEG and audio processing in general.

What's next for AudiowaveAI

Buidling version 2.0 now with streaming of content on demand and improved ux experience. Looking for more feedback

Built With

• gemini
• next.js
• openai
• prisma
• typescript
• vercel