Migrate monoliths without moving mountains.

Split was made to gradually introduce new technology to targeted groups of people—perfect for moving from a monolithic stack to a diverse group of microservices.

Start your free trial

Technologies in transit.

Many companies find themselves in a moment of transition - specialized teams, unique needs and the desire to try niche technologies are moving them away from monolithic all-in-one solutions toward a different answer, known as microservices. With the approach’s advantages in agility and the decoupling of code and processes, the value of moving to microservices is clear to engineering leaders. But the prospect of breaking apart the monolith without the infrastructure to do this in a gradual, controlled, and safe manner can be intimidating. That’s where Split steps in.

Split moves you to microservices without sacrificing user experience.

Split was made to introduce and test new technology in slow, gradual roll-outs targeted to specific groups of people. Usually we think of that as new features that users and customers will notice, but it excels at rolling out features your customers shouldn’t notice.

With Split, you can migrate features out of the monolith, one at a time, gradually moving traffic for that feature from the old monolith to the new microservice. You get complete control in this process: move the traffic from your internal employees only so you can test your microservice in production without risking your customers’ experience, or move traffic from a percentage or dimension of your users, like those in a specific location.

As you gradually move 100% of traffic to the new microservice, Split gives you the safety to switch back if at any point during the migration the microservice performance degrades. With the click of a button you can redirect all traffic back to the monolith: no need to change DNS or a Chef, Puppet, or Ansible configuration.

"Split's controlled release capability provides us a non-disruptive and dynamic way to ship features for testing and deployment. More importantly, it provides the data we need to quickly iterate on our product features in an informed manner." Qingqing Ouyang, SVP Engineering

Join the teams going beyond feature flags with Split.