Over the last several months we've been building Version 1 of our AEC-specific video transcription model for Synthesis native videos, the first project in our Community AI program. Last week, after several rounds of internal alpha testing, we released this model to a limited number of clients as part of the Synthesis Video 2 beta.
We created this Behind the Scenes video to explain how:
Video captions + transcripts will provide accessibility, search, and ultimately LMS benefits to Synthesis users.
Knowledge Architecture’s AEC-specific video transcription model performs compared to the generic video transcription model from Microsoft. (Hint — quite well!)
Our step-by-step process for building and testing the AEC-specific video transcription model works.
We protect confidentiality and intellectual property through de-identification and data masking.
Terms and sentences we have extracted from firms participating in our Community AI program to train our model have been highly valuable.
We hope this video leaves you feeling more knowledgeable and excited about Knowledge Architecture’s approach to AI and our Community AI program.
Please feel free to reach out with any questions.
About Community AI
Community AI is an opt-in program that will empower Knowledge Architecture to create AEC-specific AI Infrastructure by training AI models on client content for the mutual benefit of participating firms, while also protecting confidentiality and intellectual property.
By default, Knowledge Architecture clients are opted out of the Community program and will use the generic video transcription model from Microsoft. Clients must opt in to get access to our AEC-specific video transcription model.
So far, about 60 Knowledge Architecture clients have joined our Community AI program and contributed towards the building of our AEC-specific video transcription model.
Join the Community AI Program
Once your team is ready to move forward, you can sign up here: