# Confidence

Kernel scores its own work. Every step in the cleaning and enrichment process carries a **confidence level** - a quick way to see how certain the system is about its output.

* **High** - strong, consistent evidence across multiple sources.
* **Medium** - reasonable evidence with minor gaps or conflicts.
* **Low** - weak, conflicting, or insufficient data to be certain.

Kernel evaluates confidence across multiple dimensions of its analysis, including identity resolution, corporate hierarchy classification, and enrichment mapping. The overall confidence for an account reflects the weakest area - if any single dimension is uncertain, the account's confidence is downgraded accordingly.

Kernel also rolls these into one **overall confidence rating** per account. If an account ends up with Low confidence, it is automatically flagged for **Review**, ensuring a human reviews the record before any action is taken.

### What affects confidence

* **Strength of evidence** - how clearly available sources support the conclusion
* **Cross-source consistency** - whether multiple data points align on the same answer
* **Data completeness** - how much relevant information was available to work with
* **Internal consistency** - how well the result fits established patterns and prior values
* **User feedback** - past confirmations or corrections that reinforce or challenge the output

### Configuration

Confidence checks can be selectively enabled or disabled within the Kernel app, allowing you to control which dimensions are factored into the overall score.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.kernel.ai/confidence.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.