How it works

The platform is simple to operate: a single Unison Cloud cluster can run arbitrary Unison services and programs without needing redeploys. There's also no need to maintain separate clusters just to avoid dependency conflicts.

To create a Unison Cloud cluster, we'll provide you with a container and some environment variables to set when launching that container. In your VPC or on-prem, start as many instances of that container as you'd like to obtain a Unison Cloud cluster anywhere in the world, in minutes. Feel free to scale up or down the number of instances of this container dynamically.

All data stays with you; we never see your data or the HTTP requests sent to deployed services.

We only operate a lightweight multi-tenant control plane for managing these Unison Cloud clusters, but without access to any of the data inside them. Not only is this better for security (your data is your business, not ours), it also avoids the outgoing bandwidth costs cloud providers charge for data exiting your VPC.

Use cases

Service mesh

Support fast, ~100 µs, typed service calls and instant deployments.

More about
Adaptive Service Graph Compression Learn how it works

Distributed stream processing

Analytics, streaming ETL, or async services, as an alternative to Flink, Kafka Streams, or manual services attached to Kafka / Kinesis.

More about the library @systemfw/volturno

A function call away

Use cases are generally expressed with tiny Unison libraries (often with 10x-100x less code), and share a common communication protocol and type system.

Explore libraries on Unison Share

Try Unison Cloud on your own infrastructure

Sign up for free