Focus on feature flag data streaming

Split Software, the leading feature delivery platform, has introduced new capabilities designed to accelerate the adoption of feature flags in large-scale organizations. Feature Flag Data Streaming ensures changes to feature flags or experiments are reflected in milliseconds, keeping complex enterprise applications in sync and preventing any negative impact on user experience.

Feature Flag Data Streaming near-instantaneously propagates feature flag changes to all parts of an app -- front end and back end and/or multiple microservices. And, for enterprise environments requiring strict security, Split already offers a polling mode that eliminates long-lived open connections that can trigger firewalls. As the only feature delivery platform with both streaming and polling options, Split supports the widest variety of application requirements and client environments in the industry.

"Our largest customers serve up features to tens of millions of client apps across multiple services. Ensuring changes happen quickly and synchronously for these apps is critical,” said Trevor Stuart, co-founder and head of product at Split. “Our new Streaming Architecture addresses that issue by ensuring feature flag changes are made in a virtually instantaneous way, so applications and users are not affected. It’s an important addition to our enterprise-grade delivery platform that prioritizes data privacy, ensures fault tolerance, and always keeps features in sync.”

As part of the new Feature Flag Data Streaming capability, Split has also included a Data Hub that gives enterprises a single place to view, query and export all of their feature flag and event data. This simplifies common workflows such as validating that features will be released properly and triaging customer support issues that arise from new features.

The Data Hub consists of three components: Data Export, which makes it possible to easily export data for audit and compliance purposes; Live Tail, a real-time feed of all feature flag decisions and event data that makes it possible to quickly debug the rollout plan for complex applications; and Customer View, which provides the ability to see the state of every feature to which a user has access.