Saw a pretty interesting idea today: instead of exposing public HTTP endpoints to receive data… this guy used **email as the entry point**. Basically: you send an email → it triggers a pipeline → and the system processes everything behind the scenes. What stood out to me: * less exposure (no public API sitting on the internet) * simpler architecture * much lower cost But there are trade-offs: * email isn’t 100% reliable (retries, delays) * parsing can get messy fast * multi-tenant setups need to be VERY well handled to avoid data leaks In the end, the biggest takeaway isn’t really about email… it’s about how we’re often stuck thinking “everything must be an HTTP API”, when that’s not always true. would you try something like this in a real project? https://jch254.com/blog/email-driven-multi-tenant-ingestion-pipeline/