Thanks, Ben. This is Howard. So on Part 1, past the ITC, look, let's just step back and take stock of where solar has been over the last 30, 40 years, right? We first had to validate that it was reliable and technically sound. We've done that as an industry. We then had to prove that it could economically compete. We've done that to the point now where if you go to the Middle East, solar power is now $15 per megawatt hour, $0.015 per kilowatt hour. Okay? That's an unsubsidized market, free market. Nextracker has -- was the first company to be in the Middle East. We're there. We have an office. We are in a great position. We can compete in that market and when we have a differentiated value proposition that we just keep building with these acquisitions and our own internal organic efforts. And so what you're seeing is that solar power, as Dan mentioned, is the fastest-growing, most impactful new energy technology going in, in the United States and around the world. If you look past the ITC, if we're on a level playing field, the industry can compete. When you add storage to the equation, it's really an unbeatable firm power, dispatchable power combination. So we feel in our engagement with very large owner developers who -- their companies and their investors are pouring billions of dollars into them. Why? Because they have a durable value proposition that can compete is to provide energy to the fast-growing electricity markets in the U.S., and they just needed a bridge. And so we're in the middle of that bridge right now. We think we're in good stead with the OBBB. We're going to get more guidelines. But beyond that, we're -- it's durable, and it's unstoppable, and we feel really good about it. And that's from an industry and company perspective for solar power. Okay. AI and pricing, so right now, the way we're thinking about it, especially when you think about Onsight, who's out in the market and has robots and customers in 7 states in a couple of dozen sites, it's real technology being deployed and being paid for. It's done -- what we're migrating towards is more of a robot as a service model, where there's recurring revenue for those services, and it's in addition to the services that we provide today. But it's part of our whole platform development and constellation. Dan talked about tracker, which is core, and we're investing a lot in tracker. We've tripled our R&D spend in the last 3 years. A lot of that going to our core tracker technology, okay? And then we're building around that with these additional acquisitions, including the robotics.