Model Confidence Scores: Keeping Us Honest
The pride of the Betterview platform is the dozens of computer vision models that we have built to detect the presence of maladies and risk drivers on properties. While we are fully confident in our engineering team, we also recognize that some companies selling AI products conceal flaws in their technology. For insurers to trust them, an AI-powered insurtech must be radically transparent about the capabilities and limitations of their technology.
That is not a problem for us at Betterview – we trust the power of our technology and have nothing to hide. We regularly publish notes on the performance of our AI models and also feature confidence scores on our platform to show the accuracy of each model. If you have questions about how accurate our models are, you don’t have to take our word for it. Model performance updates and confidence scores are there to ensure you are as informed as possible about the tools you are using.