Heka is a unified data processing tool that can replace multiple existing tools like collectd, statsd, logstash, and riemann. It acquires data from various sources through inputs, transforms and transports the data through splitters, decoders, filters and a router. The data can then be output to external services, other Heka instances, or a dashboard. Plugins are added dynamically through sandboxes without needing to restart the service, and misbehaving plugins are automatically shut off. Heka aims to reduce the number of processes to manage by providing a single flexible and configurable tool that can be easily extended to handle different data processing needs.