I asked @JacquelineSYC19 about AI's data problem: "Moving data feels like a deceptively simple thing. But you're working with heterogeneous systems. Postgres is built differently than Snowflake. There's different data types and character lengths. There's small conversions you need to build along the way to keep the data consistent. As you build out the system, you have to make sure all data copied over is accurate. And that nothing is dropped along the way. Most systems can be 100's of terabytes or even petabytes of data. Every time you call an Uber, check into a hotel, send a message, or even open an app, that's all stored. All these edge cases keep compounding as you grow. As you accumulate data, some companies shard their databases. And this makes it even more complicated. Some companies have hundreds of databases flowing into a central data warehouse. And some companies even have multiple data warehouses. This makes simply storing and working with large amounts of data a very hard problem that we solved at @artie_labs."