KNIME Documentation

What is K-AI ​

K-AI is Knime's AI assistant. It works alongside you while you build workflows, without you having to leave the workflow editor. Describe a task in plain language, for example, building and evaluating a model to predict fraud from a CSV file, and K-AI builds the workflow for you. It also suggests nodes, explains errors, generates code, and annotates your workflow.

Workflow building ​

Describe what you want to achieve in plain language, and K-AI builds it. Give it a goal, for example, "filter rows where revenue is above 10000" or "build and evaluate a decision tree to predict fraud", and K-AI assembles the workflow, adding and configuring the relevant nodes.

Error explanations ​

Ask K-AI about a specific failing node. It suggests likely causes and next steps. For example, a type mismatch, a missing column, or an unreachable host.

Code and expression generation ​

K-AI can generate code for you from a natural-language prompt. Describe what you want, and let K-AI write the code into the editor. This functionality works in script and snippet nodes such as the Python Script, R Snippet, and Java Snippet nodes, in nodes using the Knime expression language such as the Expression node, and in nodes generating JavaScript views such as the Generic ECharts View. You can then edit the code generated by K-AI and ask to refine it.

K-AI code generation respects the input table schema: it reads the column names and types from the upstream node and references them correctly in the generated code.

Annotation generation ​

K-AI can generate a workflow annotation for a selected group of nodes. The annotation describes what the group collectively does — useful as a starting point that you then edit for precision.

Select one or more nodes, right-click, and choose Annotate with K-AI.

What K-AI does not do ​

K-AI keeps every data transformation in the workflow itself. Instead of modifying your data directly, it adds and configures nodes needed to apply the changes you ask for. The result stays a transparent, reproducible and auditable workflow.

Permissions ​

By default, K-AI asks for your permission before performing actions that access data, lead to node execution, or alter the workflow in significant ways. For example:

  • Inspect port object — shown before K-AI reads the data behind a node's output port.
  • Describe node configuration — shown before K-AI checks which settings of a node it can change and accesses their current values.
  • Update node configuration — shown before K-AI changes a node's settings.
  • Should I execute the selected node? — shown before K-AI executes a node. Choose to execute it, leave it as is, or type your own answer.

Respond to each prompt to let K-AI continue. For Allow/Deny prompts, optionally select Remember choice to apply your decision for the rest of the current workflow. To reset all remembered decisions for the current workflow, open the ··· menu in the K-AI panel and select Reset permissions for this workflow.

Turn on Auto approve, at the bottom of the K-AI panel, to let K-AI add, configure, and execute nodes, and check or change node settings, without prompting. K-AI still asks for permission before accessing your actual data, such as in the Inspect port object prompt. Auto approve is scoped to you: it applies across all your workflows, not to other users, and you can turn it off at any time.

Privacy and security ​

K-AI is powered by external LLM providers such as Open AI and Microsoft. Here is exactly what gets shared:

  • Your Chats: Anything you type into the AI Assistant is sent to the provider to generate a response.
  • Workflow Metadata: K-AI automatically sees your column names and data types. This allows it to suggest relevant logic and configurations for your specific nodes.
  • Actual Data (Optional): K-AI cannot see your actual data values unless you explicitly grant permission. K-AI prompts you before fetching data from a node in Build mode; permission is scoped to that workflow and off by default.

K-AI is not hosted by Knime. Instead, it connects directly to the OpenAI API, which processes the requests and sends back responses. For more information, see the OpenAI Platform Privacy Policy.

OpenAI acts as Knime's sub-processor in this mode. You and your authorized users must comply with OpenAI's usage policies, in addition to Knime's own terms.

If you want to set up your own LLM instead, see Bring your own model.

AI credits ​

Using K-AI consumes AI credits. Learn more in Understanding AI credits.