Retrieve knowledge with sources
The agent asks for company knowledge in the relevant room. NomOS returns an answer and its available supporting sources within the agent’s access rights. The agent can inspect a cited passage directly.
NomOS
Your AI works with your company’s knowledge and policies.
With NomOS, teams and AI agents put your company’s knowledge to work. Through MCP, agents retrieve applicable decisions, request checks on planned actions and submit new findings. Reviewed results remain available for the next task.
For teams using coding agents in their everyday work.
NomOS for AI agents
Your agents work with NomOS.
Retrieve knowledge, check policies and submit results for review. Connected agents can use these capabilities through MCP.
Read applicable decisions
Decisions, their status and relationships are available as structured records. The agent can inspect the basis for an earlier decision.
Contribute new findings
Agents submit knowledge and decision proposals for review. Responsible people decide what is adopted or becomes binding.
Check planned actions
NomOS assesses a submitted action against applicable policies and identifies required approvals. The call returns that assessment to the agent.
Coding agents + NomOS
An agent needs to export customer data. Which data may it use?
The agent retrieves the current policy from NomOS. Your team reviews the proposal. The recorded decision helps with the next task.
TaskAdd an export for customer data.Fictional example
Coding agent
In your development environment
Which project policies apply to exporting customer data?
MCP↔Agent requests context
NomOS
Approved project context
Project policy: exports must be scoped to the active organisation. Access and data scope must be checked.
Example: Export policy, version 3
01
The applicable policy is visible before the proposed change.
Coding agent
In your development environment
I submit the organisation-scope decision through MCP as a draft. Code review and merge are still pending.
MCP→Agent submits, person reviews
NomOS
Approved project context
The process owner confirms that the export stays scoped to the active organisation. The decision is recorded with export policy v3 and the reviewed case.
Approval by the responsible person
02
Technical permission replaces neither code review nor business approval.
Coding agent
In your development environment
I am adding a CSV export. Which policies and earlier decisions apply to this task?
MCP↔Agent requests context again
NomOS
Approved project context
Export policy v3 and the reviewed organisation-scope decision are available as references. The file format does not change that policy; data scope and permissions are reviewed again.
Exceptions do not automatically become rules
03
The rationale for the organisation boundary does not need to be found again. The new code and applicability still need review.
Illustrative workflow. MCP (Model Context Protocol) connects the agent to NomOS.
Technical action controls require integration of the execution path as well.
Understand integration and control levels ↗Inside NomOS
See the decision behind the answer.
Your team can use the shared knowledge directly, too. Here, NomOS answers an architecture question using an earlier decision.
A specific answer
For the OAuth server, NomOS names the solution already selected: Keycloak.
The source behind it
The answer refers to the record containing the underlying architecture decision.
A clear next step
Choosing a different solution requires a new or amended approval.
Reuse knowledge
Put your team’s work to use again.
Finding sources, understanding context and reviewing answers takes time and compute. NomOS retains reviewed results with their sources and scope, so others can build on them.
Start from earlier work
Previous results and their rationale are available for the next task. Your team checks what remains valid and applies to the new case.
Use knowledge across tools
Employees work in NomOS. Connected agents access the same knowledge from their existing tools.
Count both value and effort
Reusing knowledge can reduce repeated research and model work. The pilot compares this benefit with the effort of retrieval, review and maintenance.
Platform outlook
A shared foundation for your AI work.
NomOS connects company knowledge, policies and decisions. Our vision is to make this foundation available wherever employees and AI agents work.
Planned
Windows & macOS
NomGate
Control AI access at the desktop
A locally installed tool is planned to check connected AI access and agent actions against policies defined in NomOS, allowing or stopping them as appropriate.
Web & MCP
Ainomos
NomOS
Coordination and decision centre
Company knowledge, policies and reviewed decisions come together here. Your organisation defines what applies and who may decide.
Planned
Mobile access
NomApp
Use knowledge. Make decisions.
Access company knowledge according to your role and permissions. Receive notifications about pending cases, review their basis and approve or reject them.
How they are intended to work together
NomGate pauses an action
An agent wants to export customer data. The required approval is missing.
NomOS routes the case
The policy, data scope and rationale are presented to the responsible role for a decision.
The responsible person decides in NomApp
They receive a notification and review the case. The decision returns to NomOS; NomGate applies it to the requested action.
The reviewed decision remains in NomOS. A later case can build on it where it still applies.
What to know before you start.
Does NomOS work with our AI tools?
Coding agents can access NomOS through MCP. Employees can use NomOS web chat. We check which functions your specific tool supports during integration.
Explore integrations ↗Who defines policies and maintains decisions?
Your subject-matter owners review sources, approve decisions and keep them current. Your IT team configures access and integrations. NomOS makes this foundation available to teams and agents.
See how it works ↗How far does MCP control extend?
NomOS checks access rights and can assess submitted actions against policies. That alone does not technically stop an external action. Its execution path must be integrated for this. NomGate is planned to add control at the desktop.
Explore control levels ↗How do we run NomOS and where is data processed?
NomOS is planned for SaaS and operation on your own infrastructure. We agree hosting, model providers and data processing against your requirements before use.
Explore operating models ↗The 30-day pilot
Bring a workflow you want to improve with AI.
In our first conversation, we look at your task and your team’s tools. If NomOS fits, we agree a 30-day pilot covering two specific workflows.
- 01
Bring your task
Along with relevant sources and the people who know the workflow and make its decisions.
- 02
Configure NomOS together
Sources, policies and the integrations you need. We agree scope, a fixed price and evaluation criteria before starting.
- 03
Evaluate the result
Compare preparation time, quality, review effort and ongoing maintenance with your current approach.
Let’s discuss your use case.
Leave your contact details. We will get in touch for an initial conversation.
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.