Sign up for periodic email updates.
Now accepting design partners

The Runtime for Liquid Compute

Compute flows to wherever it’s cheapest, fastest, and smartest to run.

A scheduling brain profiles every Spring service, measures the call graph, and continuously re-packs services across nodes, regions, and providers—with zero redeploys and no changes to application code. One annotation. Your existing Spring beans. That’s it.

RSpond dashboard showing a 3-node cluster with live service distribution and per-method invocation rates LIVE · 3 NODES · 17K+ CALLS/SEC
Four capabilities

The cloud was designed to hold workloads. Liquid Compute was designed to let them flow.

One runtime collapses provider lock-in, transparent distribution, live rebalancing, and continuous cost optimization into a single control plane.

01

Commoditize Providers

Any premise, one cluster.

AWS, GCP, Azure, and on-prem share a single routing table. When a provider degrades or a regulation binds data, the brain treats it as a hard constraint. Services flow elsewhere.

02

Distribute Transparently

Annotate. That’s it.

The framework routes every method call—local or remote, same semantics either way. No sidecar, no control plane, no protocol to learn.

03

Rebalance Dynamically

Services move. Nothing restarts.

The routing table rewrites at runtime. New nodes activate before old ones drain. Zero downtime, zero redeploys, zero team coordination.

04

Optimize Cost First

FinOps, continuously.

Profiled services packed like bins—spot prices and call graphs are inputs. Cost optimization stops being a quarterly exercise.

Developer experience

One annotation. Your existing Spring beans.

Annotate a Java interface with @Liquid and implement it as a normal Spring bean. No base classes, no framework APIs, no protocol to learn. Same code runs on one node or across a multi-cloud cluster.

“The framework provides routing, serialization, tracing, metrics, live rebalancing, and the dashboard—without a single line of change to the impl.”

PricingService.java
@Liquid public interface PricingService { Price calculatePrice(Order order); } @Component public class PricingServiceImpl implements PricingService { public Price calculatePrice(Order order) { // business logic — no awareness of distribution var base = catalog.lookup(order.itemId()); var discount = promotions.apply(order.customerId(), base); return new Price(discount, order.currency()); } }
The consolidation play

Each incumbent owns one budget line. RSpond collapses four of them.

We’re not adding a fifth tool to your platform engineering stack. We’re replacing the four you already pay for.

Instead of… RSpond…
FinOps Kubecost, CloudHealth Cuts spend dynamically. Doesn’t just report it.
Autoscaling Karpenter, HPA Routes at the method level, not the container.
Service Mesh Istio, Linkerd Moves services at runtime, not just traffic.
Distributed Runtime Akka, Dapr Works with one annotation. No framework rewrite.
One platform. One invoice. Four replaced line items.
Live from the cluster

Every service. Every call. Every move.

Real-time visibility across the call graph, per-method invocation rates, end-to-end distributed traces, and a full audit of every rebalance.

Service call topology graph across three nodes with per-edge latency annotations
Service TopologyCall graph · per-edge latency
Grafana dashboard showing invocation counts, transport timing, latency distribution, and bandwidth usage
MetricsGrafana · Micrometer
Distributed traces in Grafana Tempo showing zero-error rebalancing across nodes
TracesOpenTelemetry · Tempo
3
Nodes
17K+
Calls / second
0
Errors during rebalance
0
Redeploys required
Why now

Three forces converging—and the timing isn’t theoretical.

AI is detonating service count, containers are hitting their architectural ceiling, and platform teams are openly shopping for what comes next.

FORCE 01

AI coding detonates.

Code is suddenly cheap and plentiful. Enterprises will ship 10x more services—and spend 10x more on compute to run them. Every efficiency point compounds.

FORCE 02

Containers show their age.

Docker is 13. Kubernetes is 12. Both were designed for the cloud era—static footprints, one region, one vendor. Platform teams are openly shopping for what’s next.

FORCE 03

New apps need new ops.

Current infrastructure can’t manage workloads interacting nonlinearly at this granularity. The tools that worked for a dozen services don’t work for a hundred.

Team

Built by engineers who have shipped enterprise systems for decades.

This isn’t a startup guessing at enterprise problems. It’s the team that built the protocols, projects, and platforms the enterprise already runs on.

Christopher Aamold

Co-founder & CEO

16 years across billion-dollar telecom and startups. Led one of Texas’s largest data transformation programs.

Matt Pavlovich

Co-founder & CTO

Apache ActiveMQ PMC. Led JMS 2.0 / 3.1. Has run the largest known ActiveMQ deployments in production.

Jonathan Locke

Principal Engineer

Windows at Microsoft. Java at Sun. Originator of two top-level Apache projects, including Wicket.

Miko Matsumura

Co-founder & Advisor

Sun’s first Chief Java Evangelist. $50M+ raised for Gradle, Hazelcast, and other developer platforms.

Get started

Built for platform engineering teams running Java/Spring at scale.

Free to adopt, paid to scale. We’re working with a small group of design partners shaping the runtime now—before pricing tiers, before public launch.