Q Could, could you talk about, um, some of the architectural decisions that enable you to do all of this and be, you know, more than this point-to-point solution? Um, like I'm thinking of like, you know, intermediate stream, processing in stream, like all the things.
A Yeah, definitely. Um, so, you know, one of the interesting things is, is that our data plan, which handles all the data movement, all the orchestration, is all open source, or will be open source components, right? Like, that's not where the magic is, right? Like Kafka's existed forever. Debezium's existed forever. Um, you know, you've had all these different components that have existed. Um, it's kind of like, you know, uh, uh, the example I gave somebody yesterday is like, Everybody shops at Whole Foods. Some people choose to make a McDonald's hamburger. Some people choose to make a Michelin five-star meal. We're the latter versus the former, right? And so we've just chosen the right ingredients to help people achieve, uh, achieve an objective, which is, you know, we have a, a change data capture service that we call Siphon, which is basically based on Debezium. But at the end of the day, like, you know, I could probably talk to a bunch of these attendees here that have tried to run Debezium at, at scale. And they run into all types of gnarly issues, right? And so we've abstracted that away, so all you have to do is just point your database, point us to your data source, uh, and we will figure out how to get data out of that in real time. Um, so that's the first part. Like, you don't have to write a, a upfront schema that says, like, hey, this is what a, a user order event lo…
AI assessment note: “we actually embed the schema into that stream”