I expect to have to answer that question a lot! Hookdeck is an event gateway, an unopinionated event log/message bus that operates over HTTP. It can be used to send webhooks, but 80%+ of use cases are for inbound webhooks (consumer side).
We think outbound is best served with an opinionated, purpose-built product, as the use case is very specific. The common feedback we got from event producers is that they are all annoyed by the complexity and costs of their current solution for sending webhooks. We think OSS / self-hosted is the solution to that. We drew from our experience handling 100 billion events, but also kept the scope to the table stakes to be highly efficient and simple to operate.
Event destinations' support is also crucial here because it means more efficient delivery with fewer errors, which can drastically reduce the overhead of event delivery.
Congrats on the launch! This is really exciting and seems like an obvious progression on how companies can expose their event streams. Interfaces like kafka are so ubiquitous compared to webhooks. This seems to build on that with infrastructure to take care of all the failure points and tough devex. Anyone who's worked with webhooks knows how common failures are!
Looking forward to getting your feedback. We feel we've got a good baseline of features with a few key ones planned (Amazon EventBridge, GCP Pub/Sub, and Kafka destinations). But we need the additional feedback to drive the roadmap.
Looks great! how is this different from your commercial offering (Hookdeck)?
I expect to have to answer that question a lot! Hookdeck is an event gateway, an unopinionated event log/message bus that operates over HTTP. It can be used to send webhooks, but 80%+ of use cases are for inbound webhooks (consumer side).
We think outbound is best served with an opinionated, purpose-built product, as the use case is very specific. The common feedback we got from event producers is that they are all annoyed by the complexity and costs of their current solution for sending webhooks. We think OSS / self-hosted is the solution to that. We drew from our experience handling 100 billion events, but also kept the scope to the table stakes to be highly efficient and simple to operate.
Event destinations' support is also crucial here because it means more efficient delivery with fewer errors, which can drastically reduce the overhead of event delivery.
Good answer.
Congrats on the launch! This is really exciting and seems like an obvious progression on how companies can expose their event streams. Interfaces like kafka are so ubiquitous compared to webhooks. This seems to build on that with infrastructure to take care of all the failure points and tough devex. Anyone who's worked with webhooks knows how common failures are!
A big +1 on the Kafka destination type being a priority. Thanks!
Really excited to try Outpost — especially appreciate that it’s open source. Looking forward to testing it in a real-world setup and sharing feedback!
Looking forward to getting your feedback. We feel we've got a good baseline of features with a few key ones planned (Amazon EventBridge, GCP Pub/Sub, and Kafka destinations). But we need the additional feedback to drive the roadmap.
Well done guys!