SoundStash

the problem

My family, my friends, and I were always swapping TikToks back and forth for the audio. Someone would find a sound, send it to the group, we’d all want to save it — and TikTok itself makes that genuinely painful. The native save flow isn’t built for sound collectors. We needed something tiny and personal: paste a link, get the audio, keep it.

what i built

SoundStash is a quiet little audio library. You paste a TikTok URL, the app pulls the sound out, and it lives in your stash. You can organize sounds into playlists, scroll your recently added, and just hit play. That’s it. No feed, no algorithm, no posting back. Just sounds you want to keep.

The empty state reads:

No sounds yet. Paste a link above to save one.

Which honestly tells you everything about how this thing works.

how it works

The frontend is Next.js. The link goes to a small backend that runs yt-dlp to pull the audio. Once it’s down, the sound gets stored and surfaced in the UI for playback, tagging, and playlists.

Most of the work was wiring those pieces together cleanly — paste box, extract pipeline, playback, list views, empty states. Standard CRUD energy, with a couple of fun edges.

tech & tradeoffs

• Frontend: Next.js
• Audio extraction: yt-dlp on the backend

what i learned

Honestly? The build itself was pretty straightforward. The hard part was what happens after the build. If you closed the app and came back later, sounds in your stash would break — stale refs, expired URLs, the usual. The fix was an automatic refresh-on-load that quietly re-validates everything before you see it. The user doesn’t have to know it’s happening. From their seat, it just works.

That was the real lesson: a tool you use every day has to feel like nothing ever went wrong, even when something definitely did.

try it

→ visit soundstash