Splash Image Inside Reflex

Reflex is high-performance data processing system whose low-latency core is structured as follows:

Each CPU in your system may run the intelligent scheduling unit, allowing you to wield the full processing power of your machines.

Each scheduling unit runs a set of simple event-driven green threads of your algorithms. The scheduling unit quickly and intelligently interleaves the various inputs of your algorithm with almost no overhead and without the need for locks.

The inputs which may drive your algorithm can be any or all of the following: the progression of time, received network transmissions, and/or outputs from other algorithms. Your algorithm code simply processes these inputs synchronously as it observes them — the inputs being delivered to you as events via callback invocations.

As your algorithms progress due to these inputs, they may choose to produce outputs themselves. These outputs are also intelligently scheduled as they are disseminated downstream out over the network or horizontally to other algorithms in the system.

Reflex Diagram

Reflex is a powerful technology product. Why does it deliver faster and more seamlessly than legacy data platforms?

The Technological Breakthrough