Find answers to all your queries on Content Lab

1: How does Content Lab work? 

Goldcast has developed a 3-step process using GPT to generate key moments and summaries from recordings and transcripts. We use LLMs and AI algorithms to generate: 

• Short video clips from your recordings based on relevance to the overall content.
• Text summaries for different use cases, like LinkedIn posts, follow-up emails, blog posts, key takeaways, and custom posts. 

The best part is that you can repurpose your video recordings to generate, edit and customize video clips and text summaries with just a few clicks, all within the Goldcast Studio. 

2: How much does Content Lab cost?  

Content Lab is included in the platform fee for all Golcast customers. Existing Goldcast users can also purchase Content Seats as an add-on to grant users access only to Content Lab. Additionally, we have a Forever Free plan for all new customers. Connect with our sales team (sales@goldcast.io) to discuss pricing and find the best package for your requirements. 

3: How secure is Content Lab? 

At Goldcast, we take data privacy and security seriously, and we're SOC2 Type 2 certified. We have implemented strict policies and procedures to ensure that client data is always protected, especially when it comes to the use of AI in our tools and services. 

Data Security on a Virtual Private Cloud (VPC) 

All client data is securely stored and processed within our Virtual Private Cloud (VPC). This means that client data is completely private and remains under our control at all times. We do not use any third-party servers to process or store client data. Our infrastructure is designed with the highest security standards to ensure client data is safe and protected against unauthorized access. 

No Data Sharing or Model Training 

We have a strict policy that prohibits sharing client data with open-source AI or any external platform. Client data is never used to train or improve any AI models, including those from third-party providers. We do not expose client data to external platforms like OpenAI or Anthropic, nor do we allow them to learn from client data. 

By keeping our AI models and vectorization technologies deployed in our own Azure instance, deployed behind a VPC, we ensure that client data is fully private and protected. Client data never leaves our private infrastructure, and no external provider has access to it. 

AI Models We Use 

For specific tasks within our platform, we utilize a combination of proprietary and open-source models, all deployed within our secure environment: 

• Large Language Models: We deploy OpenAI’s gpt-4o, and gpt-3.5-turbo in our Azure Instance 
• Text Embedding: We use OpenAI’s Ada Small vectorization model for text embedding, which we store in a postgres instance running pgvector. 
• Speech-to-Text: We use Deepgram Nova 2 and OpenAI Whisper (open source) models for converting speech to text. 
• Video Upscaling: To enhance video quality, we use the open-source GFP-GAN model. 
• Facial Recognition: We rely on the Facenet model for facial recognition tasks.

4: How does Content Lab use AI to generate video clips/text summaries videos? 

Goldcast AI generates clips/text summaries based on context and relevance to the overall video.