Get started with NomOS
In a 30-day pilot, your team uses AI with its own sources, policies and approvals. We test which earlier work can be reused in a second comparable case. You receive a joint evaluation of value, quality and maintenance effort. We agree scope and a fixed price before starting.
You identify two recurring workflows, relevant sources and a business and technical contact. We record current preparation time and review effort.
Together, we define applicable sources, policies, owners and required integrations. File import and ongoing synchronisation are different services; we agree the scope explicitly.
We compare preparation, quality, rework and ongoing maintenance with the baseline. The evaluation informs your decision about further adoption.
Both options provide the same AI workspace and impact graph. Together, we decide where NomOS runs, which models you connect and who handles each part of the operation.
From October 2026, NomOS enters production use with selected customers. Ainomos supports the rollout for both SaaS and Self-Hosted. General availability (GA) follows in March 2027.
Operated by Ainomos
For organisations that want to use NomOS and have us manage the platform.
Price on request
Production launch with selected customers from October 2026
Discuss SaaSOperated by your organisation
For organisations that want to run NomOS in their own infrastructure and take responsibility for its operation.
Individual configuration, price on request
Production launch with selected customers from October 2026
Discuss Self-HostedWeb chat and MCP integration use the same knowledge base. Agents can retrieve knowledge, submit proposals and request checks on planned actions.
Sources, policies, decisions and corrections stay connected. Technical controls depend on the integration.
SaaS does not tie you to one model provider. Self-Hosted can also use external models. Compatibility and your data-processing requirements determine the choice.
Regular curation of the impact graph requires a connected large language model (LLM). We define the model and its deployment for your Dreaming process.
For the pilot, we agree the NomOS capabilities and integrations your team needs. The planned NomGate and NomApp extensions belong to the platform outlook and are not prerequisites for getting started through MCP.
Explore MCP capabilities ↗Dreaming uses approved company knowledge to propose updates to the impact graph. People review and approve changes and remain responsible. A locally hosted LLM processes this content within the infrastructure chosen for it. This can be combined with either SaaS or Self-Hosted.
We assess the model against your content, languages and quality requirements, along with interface compatibility, model licensing and hardware needs.
We support your first 30 days with a named contact and a weekly meeting. Business Support is available in German and English, Monday to Friday, 09:00 to 18:00 Swiss time, excluding public holidays in Zug.
For critical production outages with no practical workaround, we respond within two support hours. For business-critical use, additional round-the-clock on-call support can be agreed, with an initial response within one hour for critical production outages.
Support scope and response timesAinomos with SaaS, or your organisation with Self-Hosted.
Compatible external or locally hosted models can be used with either option. For Dreaming, we recommend the controlled local environment.
In a 30-day pilot, we use two real workflows to identify the sources, models, interfaces and control levels you need. This establishes the right configuration.
An enquiry is non-binding. In the first conversation, we discuss your task and whether a pilot makes sense.