What are the costs to implement, integrate, and possibly switch from existing systems? Also, given that most sales are to the government and involve a long sales cycle, it seems really difficult for a new player to displace Modaxo solutions. Have there been any cases of displacement in the past, or has it never happened?

There are instances of displacement. It's not a nascent market, so every time we win a new deal, we are usually displacing someone. Not always, but most of the time. We do lose customers, so there are instances where others displace us. The life of one of these applications is typically long. We use 18 years as a benchmark for implementation life. Even if after 15 to 20 years you're losing a customer, it's still a high switching cost. Customers might replace the system every 15 to 20 years, but 18 years is our benchmark for replacement.

Also, considering efficiencies gained, tasks performed by the government might require fewer seats. Thinking about the government transportation side, they have been more exposed.

This doesn't significantly boost revenue because it becomes a basic expectation, much like how people expect to Google something easily. Trip planning should be straightforward and expected. Thus, it won't drive revenue growth by 20-25%. On the core side, AI won't replace our proprietary data structures, hardware integration on vehicles, asset replacement planning, and capital investment strategies. I don't see a significant risk of AI replacing our core components, except in end-user interactions, which will include an AI component.