Yeah. Sure, Najeeb. So, what we do is that we have access to a lot of data, and we actually build, if you like, a warehouse of metadata, and then we run our algorithms based on that data that's coming from thousands of dealers, and our installations are across the globe coming from different markets. So, what we are able to determine are behavior and trends and patterns on, for example, what type of products are sold in one market as opposed to another, credit risk, credit underwriting, residual values on cars, how they're moving. So that data is really a treasure trove in terms of -- if you like, we run about $300 billion worth of assets on our portfolios across the globe. So, you can imagine the amount of data that we get. And on that data, we can build large language models, which is -- primarily can be used for generative AI, and then that is how we actually build our algorithms based on data that can generate a two-way conversation with any consumer going directly on our platforms, as well as our dealers who are accessing systems are able to understand somebody's credit risk pretty quickly based on how that data presents itself. So really, we are at the -- literally the tip of the iceberg, if you like, in terms of exploiting and manipulating that data to just add so much more value into our tech stack. And going forward, you will be hearing a lot more about more specialized modules and more discrete modules that we can deploy just for AI based on the amount of data we have and how much information we have on behavioral patterns to credit risk and, in fact, down to even what type of cars are being sold, down to what color, which markets. So, really it's a very exciting time for us in terms of getting into the AI, if you like, generation and iteration to now building more use cases out of the data that we have. Hopefully, that answers your question.