AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
How to Firewall an AI Model
How does the firewall know if what's coming out of the AI model is what you want? Or if it's actually indicative of a failure or some sort? Yeah, that's a great question. So we see ourselves as like protecting the model from anything that can cause it to make a mistake. And so effectively it's firewalling off, let's say unchecked or potentially erroneous or failed AI predictions from getting out into the production environment.