# What is evidence?

Rainbird can provide evidence for every decision made. This transparency provides explainability and traceability, which ensures important decisions can be automated, trusted and audited.

We have a range of methods available to view evidence data visually, or for programmatic access via an API. 

Evidence can also be provided from different perspectives. When combined this provides a 360 degree view of decisions made.

This includes:

{% tabs %}
{% tab title="Interaction data" %}
This is a log of interactions with the engine, including any questions and answers going back and forth between the end-user and Rainbird. This provides an external view of what data was provided to the model, and the decision(s) made.\
\
For Subject Matter Experts (SMEs) this can often be enough to determine whether the decisions provided aligns to their own judgement, based on the input data. If further introspection is required, then the evidence for each decision can also be accessed.\
\
Interaction History can be accessed from the Studio under [Stats & Reporting](/rainbird/knowledge-modelling/stats-and-reporting.md). The data can also be accessed via API from the [/interactions ](https://rainbird.redoc.ly/#operation/interactions)end-point.
{% endtab %}

{% tab title="Evidence data" %}
For each individual decision made, a chain of reasoning (what we call the Evidence Tree) can be accessed. This shows what inferences were made, how they were made, the impact of the data to the certainty and the source of this data. \
\
The Evidence Tree can be viewed as a visual representation in our evidence application. This can be accessed from Studio anytime a result is shown with the (i) symbol next to it. It is also available via API from the [/evidence](https://rainbird.redoc.ly/#operation/evidence) end-point.
{% endtab %}

{% tab title="Session data" %}
This provides both metadata about the session (such as the name, ID and version of the knowledge map) but it also provides the ability to access all facts from the session.\
\
This includes all facts from all sources (hard-coded in the map, retrieved from a datasource, injected into the session, inferred from a rule or answered by a user) regardless of whether they contributed to a decision or not.

Even when data hasn't directly contributed to a decision, it can be useful to include in reports following a decision, as it can provide helpful context.\
\
This is available as an API only and provides the easiest way to extract all data from a session in order to compile reports.

For a reminder on sessions, head to [What is a session?](/rainbird/knowledge-modelling/what-is-a-query.md#what-is-a-session).
{% endtab %}
{% endtabs %}

To ensure we store or delete this data in accordance with your industry or organisational requirements a data retention policy can be set. To find out more, check out [Managing your data](/rainbird/knowledge-modelling/managing-your-data.md).


---

# 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.rainbird.ai/rainbird/evidence/what-is-evidence.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.