Product Documentation

What Ketel is for

Ketel is an AI assistant being developed for teams working with complex supply chain systems. It is designed to help teams ask questions about operational documentation, investigate how processes relate, and explore the possible effects of a proposed change.

The chat interface combines natural-language questions with answers and source citations. The usefulness of an answer depends on the sources available in the workspace and the quality of their information.

Access and sources

For a demonstration or to discuss future access, contact Ketel through the demo form. A demo request does not create an application account.

If Ketel has arranged a test session with you, use the access instructions supplied for that session. Confirm the workspace, permitted test data, and enabled capabilities with your Ketel contact before beginning. There is no public application sign-in link in this guide.

Source connections and document availability depend on the test setup. Do not assume that a system mentioned on the marketing site is connected to your workspace or that all of your organization’s information is searchable.

Ask a focused question

1. Open the chat in your authorized test workspace.
2. Describe what you want to understand. Include the relevant process, system, timeframe, and document or configuration name when known.
3. Send the question using the chat control or Enter. Use Shift+Enter for a new line.
4. Read the answer and any clarification or limitation it provides. Ask a follow-up if the scope is unclear.

For example: “Based on the replenishment procedure, what conditions put an order on hold?” Or: “What dependencies are documented for changing the minimum inventory threshold?” These are illustrative questions; answers require relevant sources in the workspace.

Review the evidence

When an answer includes citations, select a numbered citation or an item in its Sources list to inspect the supporting material. Compare the cited passage with the claim in the answer, and check that it applies to the system and timeframe you asked about.

Where available, the source inspector offers a link to the original source. Access to that source may require its own permissions. A missing or inaccessible source should be treated as a limitation, not as confirmation of the answer.

AI answers can be incomplete or incorrect. A citation helps you examine the evidence; it does not guarantee that the answer is correct or that the source is current. Confirm important conclusions with the source owner before making an operational change.

Limits and troubleshooting

• No useful answer: narrow the question and identify the relevant source. Ask your Ketel contact whether that source is included in the test setup.
• A source will not open: check that you have the appropriate source-system access. Contact support instead of sharing someone else’s credentials.
• An answer stops or shows an error: retain the error text and approximate time, remove sensitive details, and contact support if the problem persists.
• A capability is unavailable: test configurations can differ. This guide does not establish availability of any integration, automated action, export, or public API.

Safe use and support

Use only information you are authorized to share and that Ketel has approved for the test session. Do not paste passwords, authentication codes, API keys, or other credentials into chat. Treat generated answers as material to review before acting.

For help, visit Customer Support or email support@ketel.ai. Include a brief description and the time of the issue; remove confidential data from screenshots and messages.

See our Security information, Privacy Policy, and Terms of Service for related information.