Building a ENTERPRISE Speech-to-text Annotation Platform

Ask us anything
1

1) What It Is

We’re looking to build an in-house Speech-to-Text Annotation Platform. This platform aims to facilitate the conversion of audio speech into transcribed text, which can be annotated, segmented, and labeled. This is somewhat similar to existing platforms like Labelbox, Annotation Pro, AppTek, and Label Studio. I have added an image as an example as well.

http://annotationpro.org/

audio-speech-annotation-segmentation-transcription
audio-speech-annotation-segmentation-transcription1422×792 66.4 KB

2) How It Works

Basic Workflow

  1. Upload Audio: Users upload audio files into the system.
  2. Transcription: The audio is automatically transcribed into text (or manually transcribed by human operators).
  3. Annotation Interface: The transcribed text is presented in an interactive UI, where users can annotate, segment, and label the text.
  4. Review and Approval: Annotations are reviewed and approved by managers or subject matter experts.
  5. Data Export: The annotated data is then exported for further analysis or machine learning training.

3) Our Needs

User Management

  • Ability to manage hundreds of users with different roles (annotator, reviewer, admin, etc.)

Process Management

  • Workflow management to handle the stages from audio upload to data export.
  • Real-time tracking of annotation progress and quality.

Annotation/Text/Labeling Tool

  • Text-based UI for annotation, complete with features like tagging, segmentation, and labeling.
  • Capability for multiple users to work on the same project simultaneously.

Customization & Scalability

  • The platform should be customizable to adapt to different annotation guidelines and scalable to accommodate a growing number of users and data.

We’re wondering if your no-code platform can accommodate these needs and functionalities. Looking forward to your insights!

2

Yes.

1 Like
3

Yes. But.

  1. It’s not just about WeWeb, it’s also about the backend you choose.
  2. You might hit some limits where you may need some custom code, this one depends on how flexible you are.

Would love to see this happen on WeWeb :slight_smile:
As of now, I think WeWeb is the underrated king in the no-code field. Especially in terms of how much you can bend and shape it.

Edit: Personally I’d even go so far to back this up, as say that I’d even put money in WeWeb, if there was a chance, because I feel like it’s gonna explode and be the next (better) Bubble soon.
@Slavo #takemymoney

2 Likes
4

Thanks @Slavo @Broberto .

Well, it’s a good start to hear that it’s possible. I would like to understand how you would go about creating the UI component as shown above. Is this something that needs custom code, any librabry/plugins possible to use - or can it be created entirly from weweb elements?

5

What would be the exact function of it? I see some audio wave divided in sections

6

Hey Erik,
Creating the UI component demonstrated above often involves a combination of approaches. Utilizing custom code, available libraries/plugins, and weaving together web elements within Weweb might be necessary. It’s often a blend of these resources to achieve the desired UI outcome. You can get help from some text to speech tools too.
Hope this helps :grinning: