On the second question, Mizrahi, yes, we have been working a lot with AI on different verticals here. I will mention some of them. The first one is internal productivity. It can be for engineers. It can be for management people. So we have been working on operations, on customer experience. So we have a lot of use case live right now more on the productivity side. The second one, we have been trying -- I would say, the idea here is not like to replace advisers, but how we enhance our advisers using different agents. It can be a relationship agent, AI agents. Of course, it can be transactional agents. So we have been creating a lot of this type of agents to give more productivity and to increase the level of service that we deliver to our customers. And I would say a third one, we haven't been investing a lot on portfolio allocation of our customers. So we have been creating more rules. We have been creating a centralized portfolio allocation, and we are using a lot of technology to make it happen. So there are many use case here, some already at scale, some at very early stage. Another one that -- it's already at scale. Today, we listen, read, hear everything that our advisers, especially internal advisers today, 100% of them. And we can classify all the conversations, all the interactions with our customers. It's a product offering, if it's only a relationship activity. So the level of information and sales management that we have today, it's very, very high and we will continue to invest on that. But again, not to replace advisers, okay, but to enhance them. Of course, when you go to very small clients, more like a digital or what we call here self-direct clients, then you can have like a fully deployed AI solution, but that's a very small part of our business today. It could grow in the future. But again, the focus here is how we enhance the advisers and how we free them like to focus 100% of their time on relationship, not like on operations or allocation or other activities that don't generate value for our customers.