So from the look of this, you’re open sourcing all of the agent/usage code, but I’m guessing you’re keeping the trained model in-house, so that becomes the valu

So from the look of this, you’re open sourcing all of the agent/usage code, but I’m guessing you’re keeping the trained model in-house, so that becomes the value prop for Dataherald the product - the trained LLM?