Orama Lake, open source

Own your AI

An AI easy button, with batteries included. Everything you need to design, train, and ship real AI that learned on your own data, on your own hardware. MIT-licensed, all of it.

Released

Orama Lake v0.1.1

Get the source, with every component pinned at versions that work together:

git clone --recurse-submodules https://github.com/oramasearch/lake-platform.git

Read the release notes

What it is

An AI platform for your whole team

Orama Lake is an AI platform. It lets you shift left, and makes it easier for your whole team to create and ship apps, agents and AI. You use AI to create and ship AI.

It's the same code that runs lake.live, and it's yours to build on.

From AI research to production in the same day

Experiment tracking, a streaming agent substrate, and differential source control, builds and deploys are one system, so what works in the morning can be in front of people by the evening.

What's in the box

Four parts, and the agents and models

Lake

Where you work. Workspaces, agents, apps, repositories, search, file storage, sign-in, secrets, a container registry; one API, served as REST, OpenAPI and MCP.

Zetta

The engine, in C++26. Your data, SQL, full-text search, and machine learning with an explanation for every prediction.

liquid

Differential source control, builds, packages and deploys in one program. Every write is captured and signed.

The operator

Infrastructure as software. Runs it all on your own Kubernetes clusters, always reconciling.

Agents and models

The agent that builds agents, the one that builds apps, the one you ask, and the models they run on.

Run it where your data is

From a laptop to any Kubernetes cluster

Try it

On a laptop, on a single-node k0s or k3s.

Run it

One machine or a small cluster, with k0s and the operator.

Scale it

Any Kubernetes cluster, from the operator's helmfile.

Or let us run it: lake.live

It's yours

MIT for everything we wrote

  • MIT-licensed, all of it.
  • No enterprise edition of the code, and nothing held back.
  • Contributions come in under a Contributor License Agreement.
  • The names and logos are trademarks; a fork picks its own name.

How Orama earns

The code is free and stays free

We earn on what goes on top of it.

lake.live

We run it for you.

Forward-deployed engineering

Our engineers build with your team.

Enterprise support

Support with an SLA.

A place for what you build

Where people offer what they build on the stack. Planned.

Join in

Build it with us

Issues and pull requests are open to everyone. The contributing guide says how to build it, how a change gets in, and what to sign first.

Free, sovereign AI, for everyone.