simmis build with simmis

program your organization,
not just your technology.

build with simmis join cloud waitlist
our open-source pledge

TRUSTED BY

Organizations leveraging the Simmis stack.

Stub Traceable accounting at scale 5,000+ businesses · ~98M-fact workload Arbetsförmedlingen Swedish labour-market ontology 40,000+ connected concepts · government production Heidelberg University Longitudinal research Emotion tracking across time Sonomos AI-native operations Exploring business operations on the stack

AI work needs organizational control.

not more wrappers, integrations, or fragmentation.

MANAGEMENTknows this

“The demo looked like magic. But the tool wanted clean, connected data and decisions made in consistent, repeatable ways.”

Julie Averillformer Global CIO & EVP · lululemon
lululemon The New York Times

“Greater value comes from disciplined platform-orientation, not tool-proliferation.”

Monica CaldasExecutive VP & Global CIO · Liberty Mutual
Liberty Mutual CIO

“We have established a global technology stack together with a group-wide AI governance framework.”

Fabian WinterChief Data & AI Officer · Munich Re
Munich Re Munich Re interview

“That’s why we have a governance framework and set up a control plane.”

Tan Su ShanCEO · DBS Group
DBS Group DBS interview

“All of the AI projects we have observed … are failing. Every single one – we have seen 0% success in a year and a half…”

Nikhil SureshExecutive Director · Hermit Tech
Hermit Tech Hermit Tech

WORKERSfeel this

“I have 4–8 different sessions running… managing all of them and following up is non-trivial.”

Reddit

“The job shifts from writing code to managing external memory for the agent.”

Hacker News

“I always want to be manually verifying or prompt writing, and keeping it all straight is taxing.”

Hacker News

“The new workflow amounts to reviewing someone else’s code all day. Forever.”

Reddit

“Claude’s speed means I don’t get much of a break before it’s on to the next thing.”

Reddit

LAWdemands this

MOFFATT v. AIR CANADA · 2024 BCCRT 149

Air Canada remained responsible for false information its chatbot gave a customer.

CanLII decision
EU ARTIFICIAL INTELLIGENCE ACT

The Act assigns covered providers and deployers duties around oversight, logs, monitoring, and use.

EUR-Lex

Say no to the Wild West of piecemeal AI patchwork.

Say yes to:

  • Unified organizational ontology.
  • Bounded environment for human-agents collaboration.
  • Controlled path from individual attempt to organizational change.
RESEARCH · POSSIBLE FUTURES SIMMIS · CONTROLLED ADOPTION DVERGR · BOUNDED WORK DATAHIKE · VERSIONED ONTOLOGY
01 MEMORY
Datahike Versioned knowledge substrate

Know your business as it was, as it is, and as it could be.

Datahike gives people and agents one connected, versioned memory of the organization without losing what came before.

Understand the business through one

into structured organizational memory.

See of what was behind decisions and analyses

before making them real.

GDPR-READY PRODUCTION-PROVEN CORE

27× fasterrelationship joins
<1 msto create an in-memory branch
33k+ assertionsacross 2,659 automated tests
40k+ conceptsconnected in production
5k+ businessesserved by a production system

Open source · PostgreSQL-compatible access · No read server required

build with simmis join cloud waitlist Try Datahike live ↗
02 TEAMWORK
Dvergr Bounded organizational work

Scale the work, not the coordination overhead.

Dvergr keeps each assignment’s context, participants, tools, and limits together, so people and agents can move work forward without constant reconstruction or manual routing.

of organizational ontology as agents work.

Let work .

Pause, long-running work without starting over.

Give agents real tools and budgets within .

DATAHIKE-NATIVE · BUILT FOR LONG-RUNNING WORK

ACT ON LIVE KNOWLEDGENative Datahike query and transaction inside agent work
WORK OUTLIVES THE WORKERRooms retain messages, participants and process state
HAND OFF WITHOUT REBUILDINGPeople, agents, scripts and policy share one participant model
AUTONOMY WITH RUNTIME LIMITSScoped tools, credentials, budgets and execution boundaries

OPEN SOURCE · DIRECT DATAHIKE ACCESS · PERSISTENT ROOMS · SCOPED EXECUTION

build with simmis join cloud waitlist See Dvergr in the stack ↗
03 CONTROL
Simmis Deliberate organizational change

Decide what becomes organizational truth and progress.

Simmis separates work attempts from accepted organizational state. Its open-source interface lets authorized people manage organizational records and work directly with agents in persistent multi-person, multi-agent Teams—all without writing code.

Choose a scenario. Engineering, Marketing, and Accounting are example workflows shown in the real Simmis interface, lightly enhanced for legibility.

Engineering recover a failed operation

A team diagnoses a provisioning failure, tests a proposed correction, challenges its recovery design, and accepts the reviewed result.

Ready
  1. 00:00–00:23.6 · Define the failure

    A person defines the incident, constraints, and authority for a shared retained Team context.

  2. 00:23.6–00:42.5 · Inspect and test

    Agents inspect an imported source snapshot, reproduce the failure, test the proposed operation, and expose their tool evidence while accepted state remains unchanged.

  3. 00:42.5–00:57.5 · Review the proposal

    The semantic summary orients the reviewer while the complete code and knowledge diff remains available for exact review.

  4. 00:57.5–01:03.5 · Retain the decision

    The accepted recovery and identity decision becomes durable organizational memory linked to its proposal and evidence.

  5. 01:03.5–01:16.5 · Inspect accepted history

    A question is answered from retained context, and Timeline changes only after the viewer selects a past point.

  6. 01:16.5–01:30 · Schedule controlled recurrence

    The recurring check produces regression evidence and a review task on failure without merging a change automatically.

see how the stack carries state and change

build with simmis join cloud waitlist
04 FORESIGHT
Simmis Research Possible organizational futures

Compare futures before choosing one.

Simmis is exploring how people and agents could compare possible futures from retained evidence before an authorized decision is made.

EXPERIMENTATION ILLUSTRATIVE RESEARCH DIRECTION
What happens if we hire three senior engineers in Q3 instead of Q4?

Hiring in Q3 may reach the delivery target earlier, but increases near-term spending and onboarding pressure.

DELIVERY TARGET Q3 HIRING OPTION CURRENT Q4 PLAN
01 · EVIDENCEONE RETAINED BASIS
STAFFING SPENDING DELIVERY DECISIONS
RETAINED + VERSIONED HISTORYSAME EVIDENCE FOR BOTH PLANS

VERSIONED ORGANIZATIONAL EVIDENCE · DATAHIKE

02 · SIMULATIONSAME BASIS · TWO SCENARIOS
RETAINED EVIDENCE
CURRENT Q4 PLAN
Q3 HIRING OPTION
TEST BOTH
DELIVERY TARGET SPENDING LIMIT TWO ONBOARDING SLOTS

YGGDRASIL BRANCHES · DVERGR RUNS

03 · COMPARISONPROJECTED RELEASE-READY CAPACITY
JUNJULAUGSEPOCT DELIVERY TARGET ABOUT SIX WEEKS ONBOARDING CAPACITY 3 HIRES · 2 SLOTS Q3 HIRING OPTION CURRENT Q4 PLAN

The Q3 option reaches release-ready capacity about six weeks earlier. It increases Q3 spending, while onboarding remains the limiting factor.

PROJECTED OUTCOMESSPEED · SPEND · CAPACITY
PERSON OR POLICYAUTHORIZED REVIEW
KEEP CURRENT PLAN CHOOSE Q3 OPTION

comparison overview

EVIDENCE Staffing, spending, delivery, and decisions come from retained, versioned organizational history.

SIMULATION One agent tests the current Q4 plan and Q3 hiring option against the same delivery target, spending limit, and two onboarding slots.

COMPARISON The Q3 option reaches release-ready capacity earlier, but costs more in Q3 and increases onboarding pressure.

read the vision

maximum openness. maximum trust.

SaaS normalized holding organizational memory hostage inside proprietary infrastructure. We will not repeat that mistake for AI. Simmis is open from its programmable interface down to its versioned database. Use it, inspect it, and change it on your terms. You keep your data and the code needed to read it.

01

Your data remains yours.

Access to organizational memory should not depend on a vendor account.

02

No black boxes.

Inspect the code and understand how the system handles state and change.

03

Modify any layer.

Adapt the product, replace components, or extend the stack around your organization.

04

Built to outlast us.

The source and the code required to read your data remain public.

OPEN THROUGH THE STACK
build with simmis read the licence see the architecture

Top FAQs

read all FAQs →
01What can Simmis do today?
The working alpha combines persistent Teams, shared organizational state, bounded agent work, branches, proposals, and retained history. Packaging and production guarantees are still developing.
02Why not use an agent harness, coding agent, or AI workspace?
Use whichever tools make good workers. Simmis gives them shared organizational state, contained attempts, deliberate adoption, and history that survives the worker.
03Does every agent action require review and a merge?
No. Agents can take many steps inside one bounded attempt. Review happens when proposed work is about to change accepted organizational state.
04Do I have to move my entire organization into Simmis?
No. A new organization can begin with Simmis as its operating core. An existing one can start with one consequential operation and expand as related work begins sharing state.
05Why not build this with Postgres, Git, chat, and an agent framework?
You can. The difficult part is keeping the database, repository, instructions, execution basis, and retained history coherent as they change.

put more AI to work. keep control of your organization.

build with simmis join cloud waitlist