Automating video transcription workflow

[kegsdoin]

Dear Deepgram Team,

I'm exploring solutions for an automated video transcription workflow for our house of worship archive and ongoing needs and would like to confirm if your services support the following requirements:

Automatic detection and processing of new video uploads
Synchronized video-transcript viewing capability
Searchable transcript database
Custom dictionary/vocabulary management that can be refined over time
Whisper-level transcription quality
Automated caption upload back to YouTube
Our ideal workflow would process new videos automatically, maintain searchable transcripts, and allow us to continuously improve accuracy through custom vocabulary training.

Could you please confirm which of these features are available in your solution? If some aren't currently supported, are there workarounds or planned features to address these needs?

We have roughly 1000 hours of video, perhaps 5000 hours of audio in the backlog, plus we are creating roughly 5 hours of new content per week.

Thank you for your time.

Best regards,

Kevin

[deepgram-community[bot]]

Thanks for asking your question. Please be sure to reply with as much detail as possible so the community can assist you efficiently.
_{Consider joining our Discord community for more opportunity to engage with your fellow Deepgram users. You can earn points which can be redeemed for cool stuff by being active in our communities!}

[deepgram-community[bot]]

Hey there! It looks like you haven't connected your GitHub account to your Deepgram account. You can do this at https://community.deepgram.com - being verified through this process will allow our team to help you in a much more streamlined fashion.

[deepgram-community[bot]]

It looks like we're missing some important information to help debug your issue. Would you mind providing us with the following details in a reply?

• The programming language you are working in (e.g. JavaScript, Python).
• The deepgram product you are using (e.g Speech to Text, Agent API)
• A request ID that triggered your error or issue.

[kegsdoin]

Hello, I'm not a deepgram customer yet, but would like to use Python and speech to text.

[deepgram-community[bot]]

Welcome, Kevin! To address your needs for an automated video transcription workflow, Deepgram's services can offer the following:

1. Automatic Video Processing: Deepgram provides APIs that can be integrated into your systems to automatically process new video uploads.

We support over <https://developers.deepgram.com/docs/supported-audio-formats|100 types> of audio formats so depending on which format your videos are in, we should be able to handle them. You'll want to check out our Docs on Pre-Recorded <https://developers.deepgram.com/docs/pre-recorded-audio|Speech To Text> to learn more.

2. Synchronized Video-Transcript Viewing Capability: While Deepgram delivers high-quality transcriptions, integrating this feature with your video player might require additional development to ensure synchronization.

3. Searchable Transcript Database: You can store transcripts in a database to make them searchable. This would require setting up a system to manage and search through transcripts post-transcription. Deepgram doesn't currently offer a feature like this.

4. Custom Vocabulary Management: Deepgram supports custom vocabularies which allow you to improve and refine transcription accuracy over time. There are different features you can use like:

• Custom Model training (Enterprise only)
• https://developers.deepgram.com/docs/keywords|Keywords
• <https://developers.deepgram.com/docs/find-and-replace|Find & Replace>
• https://developers.deepgram.com/docs/search|Search
• Key Terms (coming soon)
5. High-Quality Transcription: Deepgram provides high-quality transcription services that are competitive with Whisper. Checkout our <https://deepgram.com/learn/nova-2-speech-to-text-api|Nova-2 Benchmarks here>. We are also release a brand new model very soon, called Nova-3, which should be superior to Whisper.

6. Automated YouTube Caption Uploads: While Deepgram transcribes videos, the automation of caption uploads to YouTube would need to be an additional service. You could use SubRip for this. Deepgram generates JSON-based transcripts, but you could export the output to .srt using external processing or additional formatting an use that with YouTube. Synchronized transcript viewing or automated YouTube uploads would need some custom implementation on your end.

Hope this helps! If you want connect real time with others building with Deepgram <https://discord.gg/deepgram|join our Discord Community>.

This message was sent by John Vajda from Deepgram, via our community automation.