Maintain knowledge once. For people and AI agents.

A work instruction helps an employee carry out a task. The same instruction can tell an AI agent which conditions to consider. In PULSE, both use the same maintained knowledge base. A business rule is changed once, making the update available for AI use as well.

The surrounding context matters too: Which process does the instruction belong to? At which step is it needed? Which role is responsible? PULSE connects knowledge to this information, so a rule can be found and understood in relation to the task.

The need for a change often becomes apparent while reviewing a process. An approval threshold may be outdated, a responsibility may have changed, or an instruction may no longer fit a task. In PULSE, you can open and edit the relevant knowledge directly from that context.

You can also start from the knowledge overview, a linked role or a system. Each route leads to the same knowledge entry. Once it has been updated and published, the new version is available wherever that entry is linked. This reduces duplicate maintenance and makes it easier to see which tasks the change affects.

The process itself can evolve too. When steps, responsibilities or the assignment of an instruction change, you maintain those relationships in the same business context. People and agents receive both the content and its place in the process.

In our PULSE demonstration data, the sales approval threshold increases from €25,000 to €50,000 net. Up to and including €50,000, Sales reviews the quote independently. Above that amount, the quote requires approval from the Head of Sales.

1. The responsible person updates the policy in PULSE and publishes the new version.
2. The same policy remains linked to the quote approval process. The new version is now available there.
3. A connected agent retrieves the published policy and its process context for its task. The new threshold becomes part of its working context.

A €40,000 quote makes the difference clear: under the old amount rule, the Head of Sales had to approve it. Under the new rule, Sales reviews it independently. This is the change that employees and agents should be able to draw from the shared foundation.

The example illustrates the policy change using our demonstration data. A case like this is useful for practical verification because the expected result actually changes. The answer should also refer to the appropriate published version.

The walkthrough shows how the policy, process and responsible role fit together. Explore quote approval in PULSE

Keeping the shared foundation reliable requires a clear business decision about each change. A new instruction may be prepared today but only take effect next month. An older case may still be governed by the previous rule.

• Version: What has changed compared with the previous version?
• Applicability: From when, and for which cases, does the rule apply?
• Ownership: Who reviews the content, approves it and resolves open questions?

PULSE supports this work with versions, change comparisons and publication of the agreed content. The business owners record when and where that version applies. Employees and connected agents can then trace the basis for a particular case.

This is the value of the connection: when your teams keep rules, instructions and process relationships current in PULSE, they also maintain the foundation for working with AI. Knowledge is edited in a form people in the business can understand and made available to connected agents for their tasks.

An external agent is given access to the PULSE content it needs. You define which sources it may use and which tasks it should perform. When it retrieves its working context, the published version is available through that connection.

Start with a process whose rules change regularly. Link the required knowledge entries, name the responsible people and configure access for your agent. With the next change, check that the updated foundation is available in the process and reflected in the agent’s answer.