AI-powered video subtitling with open-source flexibility and smart features

In an internal initiative at Softhouse, we explored how AI-powered auto-transcription can make video content more accessible and useful. Using OpenAI’s Whisper model, we built a solution that automatically transcribes speech to text – supporting both legal compliance for video captioning and smarter access to information within video content.

This is more than just an accessibility tool. It also enables intelligent search, summarization, and internal knowledge sharing—while offering full control over both input and output, thanks to its open-source nature.

What we’ve done

• Implemented OpenAI Whisper as the core transcription engine

• Built auto-captioning support for internal videos and demos

• Evaluated potential for smart video search and summarization

• Explored how the solution enhances transparency and data privacy

Business value

• Enables legally required video accessibility through automatic subtitles

• Increases internal efficiency with AI-based video text analysis

• Enhances transparency and access to knowledge

• Provides a strong foundation for further open-source AI adoption

We code the future – with AI that listens and understands.