Yeah. I know. it is a great question, Christian. So I think I will maybe talk about that in 3 parts. Right? So first, as you mentioned, right, our approach is around the notion of helping whatever type of company uses the product that when there is more traffic more complicated traffic, and people are worried about latency, that is a good fit for our solutions, independent of you know, what exact application may be. So certainly, the growth in traffic driven in some part due to AI, which is much more, you know, burst traffic, high volume traffic, and different than what we used to see it results in 2 parts. 1 is just managing that traffic efficiently and with low latency. And second is, dealing with the new kinds of threats that arise because of the use of AI by even enterprises, but by everybody on the network. So that is certainly 1 of the drivers. And I would say we see that as, you know, kind of a sustainable thing for several years. Because it is related ultimately not to who is building how many data centers. it is linked to are people using AI? Right? And I think we all can answer that for ourselves. So that is clearly 1 of the drivers. Second, I think you talked about Microsoft and, obviously, we have been partnered with them. it is been a great customer for us for a long period of time. The expansion is that over the next couple of years time frame again, we are more and more aligned on the road map that as they build out their own architecture and infrastructure that we are continuing to be embedded with them in broader areas than we used to before. Right? So it is continuing to expand that. At the same time, obviously, we are very focused on ensuring we do our part to help them continue to be successful. So that is simply kind of a reflection and joint commitment from both sides that we want to make it work for an extended period of time, and this is just a reflection of that. And, ultimately, has to be that we have to deliver the right technology they need and we are obviously focused on that from a technical point of view. The last point you mentioned actually is an interesting 1. So I would say this is not a big demand driver immediately, but certainly in the next 2 to 3 years, as you see in the news, large enterprise, particularly chooses to do sometimes their own models, or because of sovereign AI reasons in outside of US as well. They choose to do actually more things on-prem than even in the cloud. Our solution can go across all of that. So as we see uptake in the next 2-3 years with enterprise doing more AI, like inference AI and generally the large ones preferring to do their own. We see that obviously as a strength for us, again, based on the nature of our solution. And the ability that we have maintained how we deliver these capabilities in any kind of form factor. Great. Thank you for that.