LOCAL AI. OPEN HARDWARE. YOUR DATA STAYS WITH YOU.

Digital infrastructure
for you.

somsorium brings AI, collaboration, research and open infrastructure together in one system you can keep under your own control.

Runs on your own hardware

No forced cloud

Only what you approve is allowed

Will be published under AGPL-3.0

Every action is logged

Your data stays yours

Built in Leipzig

The system

Four building blocks. One open ecosystem.

somsorium connects personal AI, communication, research and local infrastructure. Each part works on its own — together they make a system in which data, knowledge and tools can stay under your control.

From a question to shared knowledge

One algae culture. One question. Many experiments.

What happens when an experiment does not end with its result — but becomes something others can build on?

A microalgae culture grows inside a laborium cassette. We want to understand how different conditions change its growth.

A simple question becomes a documented experiment — and that experiment can become the starting point for a shared research question.

Question

How much light is enough?

Microalgae need light to grow. But more light does not necessarily mean more growth.

What changes when we vary the conditions deliberately?

Three culture vessels under different lighting Schematic: three identically built cassettes side by side, labelled “low light”, “medium” and “high light”, with a different number of light lines above each. The cultures inside differ in density. The drawing shows differences — not measurements, and not a ranking. low light medium high light
Illustrative experiment setup.
Experiment — laborium

A question becomes a controlled experiment.

laborium keeps track of what we change, what stays constant and what we observe.

The culture runs in a laborium cassette. Experimental conditions, measurements and observations stay connected to the experiment.

Experiment Aillustrative
Light cycle
12 h / 12 h
Temperature
— example value —
Medium
— example value —
Started
— example value —
Illustrative experiment record — no real lab values.
Observe

The culture changes. So does the data.

Measurements and observations accumulate over time.

Not every difference is immediately visible — and not every change means the same thing.

The same culture at four points in time Schematic: the same culture vessel on day 1, 3, 5 and 7, the culture slowly getting denser. The running thread connects the four points in time. Illustrative progression, not measured data. Day 1 Day 3 Day 5 Day 7
Illustrative progression — not measured data.
Understand — smolit

Spot patterns. Ask better questions.

smolit can help compare measurement series, surface unusual changes and explore possible relationships.

Why does one culture develop differently from another? Which condition is worth testing next?

Three measurement series side by side Illustrative analysis view: one path becomes three runs for experiments A, B and C, which come back together afterwards. No axes, no measured values. Experiment A Experiment B Experiment C

smolit · illustrationExperiment B changes more noticeably after day 4.

Illustrative analysis view.
Share — mesorium

A result does not have to end in your own folder.

Results, images and context can be shared and discussed through mesorium.

That turns an experiment into something others can see, question and build on.

Full structured publication of the experimental setup, data and analysis is one of the parts we are continuing to build.

mesorium connects publishing, community and communication.

Illustration

Microalgae · light experiment

How does light change the growth of a microalgae culture?

Results · Discussion

Discuss — mesorium

Others may notice what you have not seen yet.

A research group has seen a similar effect.

An SME contributes practical experience with cultivation or lighting.

A maker suggests a different setup.

A school class wants to recreate a simplified version.

One result becomes a discussion — and that discussion creates new questions.

Four perspectives on the same experiment Four short side branches lead from a point labelled “experiment” to research, SME, maker and education. They do not converge again. Experiment Research SME Maker Education
One observation, several perspectives — schematic.
Replicate

The same experiment. Different conditions.

Others repeat the experiment — perhaps with a different light cycle, another culture or a changed environment.

Their observations can be shared again and compared with the original.

One experiment, repeated in three variants Three paths lead from one starting point to three new culture vessels — different light cycle, different culture, different environment — and come back together afterwards. different light cycle different culture different environment
Schematic sequence — no claim about numbers or outcome.
Knowledge

An observation becomes something others can test.

One experiment can provide an indication.

When others can understand it, change it and repeat it, a more robust picture can emerge.

That is where research, AI and collaboration meet in somsorium.

Several paths coming together The paths of the repeats meet in a single calm point, labelled “shared knowledge”. shared knowledge

The smolit node can run the participating services locally.

The context

What is changing right now.

AI is moving from a chat window to a tool that can work with files, accounts, devices and workflows. At the same time, more and more of our communication, work and data sits on platforms belonging to someone else.

We are building the other direction: systems that can run locally, work together openly and stay possible to follow.

Local, where local makes sense.

Connected, where working together needs it.

Open enough to stay checkable.

Where we stand

Where we are.

Some parts are running, others are being built. We would rather tell you which is which than let you guess.

Today

  • smolit is a release candidate with a reproducible build.
  • mesorium runs for private areas and connected groups.
  • laborium runs locally and holds experiment and measurement series together.
  • For the node there is a lab device and a tested build pipeline.

Next

  • The audit view in smolit.
  • Public read access in mesorium.
  • The first complete run from question to published result.
  • The repositories open under AGPL-3.0.

After that

  • Parts list and build plan for the node.
  • The lab hardware, so others can rebuild it.

Test network

somsorium is opening step by step.

Our first test network is planned for up to 5 organisations and 500 individuals. We are not assigning places yet. If you would like to try somsorium early, you can join the waitlist now — it commits you to nothing.

Non-binding. We’ll be in touch once we open suitable places.

Leipzig

Who is behind this.

Johanna Gaßmann

Founder & managing director

If you write to us, she answers.

Sam Schimmelpfennig

Engineering

Architecture, code, hardware, systems — all four building blocks.

More about us, how we think and where this is going: About and Vision.