← Back to selected work

Caption-aware transcript tool

YT Transcriber

A browser-first utility that turns available YouTube caption tracks into searchable, timestamped text and exportable files.

ViteNode.jsYouTube captionsProduct UX

YT Transcriber pairs a focused transcript workspace with strict video validation, language selection, search, timestamp links, and TXT, SRT, and WebVTT exports. It is a public beta: availability depends on whether YouTube exposes captions to the request source, and it does not bypass private or restricted media.

RoleProduct, UX, and full-stack implementation
FocusCaption extraction, search, export, resilient errors
OutcomeA compact transcript workflow that stays clear about caption availability

THE OPERATING FLOW

01Validate a YouTube URL or video ID
02Choose an exposed caption language and search the transcript
03Jump by timestamp or export TXT, SRT, and WebVTT

CASE STUDY

The challenge

Transcript tools often hide whether a result is real, which language was selected, or why a video cannot be extracted. The product needed a fast, readable flow with honest failure states.

What I designed and built

  • Replaced placeholder success states with a same-origin Node extraction route backed by a community caption package.
  • Added language selection, transcript search, timestamp navigation, bounded responses, caching, and rate limits.
  • Kept the interface explicit about public-beta limits: caption availability varies by video and request origin.

Outcome

  • A focused workflow from video URL to searchable transcript and download-ready caption formats.
  • A reusable example of pairing a polished interface with transparent upstream reliability boundaries.

This is a public beta. It only works when a caption track is exposed to the service; it does not bypass private videos, restricted captions, or YouTube access controls.