Skip to main content

Overview

Single Self-Contained Binary

Event collection infrastructure comes in all shapes and sizes except a single artifact that can be deployed anywhere. Buz changes that.

Buz was created from the ground-up to eliminate as many moving pieces as possible without sacrificing quality and transport guarantees.

TL;DR
  • Less headaches for infrastructure humans.
  • Decreased infrastructure costs
  • Increased operational efficiencies.

Collects Events from One or More Sources

Most event collection systems are single-protocol, which results in duplicate infrastructure when more than one protocol must be collected -> webhook, snowplow, segment, cloudevents, etc.

This is not the case with Buz.

Buz supports a number of common input protocols and will continue to support more.

TL;DR
  • Single, flexible system instead of N pipelines to support N protocols.

Validates Events on the Edge

Streaming data is all about one thing: increasing the speed of action and decision-making. If events are not validated fast, decisions and actions cannot be made with the conviction they require. Who cares about making bad decisions, on bad data, fast? Nobody.

TL;DR
  • Incoming events are validated immediately.

Sends Events to One or More Destinations

Buz was purpose-built for flexibility and does not require provisioning more infrastructure to fan out events to multiple places.

Multiple Buz sinks can be configured regardless of what cloud the destination actually resides in.

TL;DR
  • Send events to one place or five. No additional infrastructure necessary.

Generates Rich Metadata

After years of building, maintaining, and managing event tracking systems there's one thing that has consistently stuck out:

More metadata and expedited knowledge of what is happening within the stream would be unbelievably empowering.

Metadata often happens at the end of the pipeline, in the data warehouse. Not awesome.

TL;DR
  • More visibility into events, faster.

Eases Operational Burden

Buz is easily configured with a single yml file. This file is self-validating and, when using the vcode yaml plugin, practically writes itself.

It can be rapidly deployed and re-configured, and scales well as needs do.