> ## Documentation Index
> Fetch the complete documentation index at: https://docs.nebuly.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Get Interactions

You need an API key to authenticate. See [here](/embedded-pages/quickstart) for more information about the API keys.

<ParamField body="time_range" type="object" required>
  The time range to consider when computing the metrics and analytics to show.

  <Expandable title="properties">
    <ParamField path="start" type="string" default="2024-05-30T09:27:52.420Z" required>
      The start date for the time-range.
    </ParamField>

    <ParamField path="end" type="string" default="2024-05-30T09:27:52.420Z" required>
      The end date for the time-range.
    </ParamField>
  </Expandable>
</ParamField>

<ParamField body="filters" type="object[]" required>
  Filters you can use to get data based only on the interactions you are interested in.

  <Expandable title="properties">
    <ParamField path="kind" type="string" required>
      The filter kind. It can be one of the following values:

      * user\_intent
      * negative\_user\_intent
      * keyword
      * user\_feedback
      * tag
      * type\_of\_problem
      * topic
      * user
      * user\_query
    </ParamField>

    <ParamField path="tag" type="string">
      Optional field to be used only when the kind is tag. It should specify the tag name, which is the same as the key you used in the tags dictionary during the interaction submission.
    </ParamField>

    <ParamField path="values" type="string[]" required>
      The values of the selected filter you want to filter on. Note that all the values passed in
      the list are considered to be in "or" the one with the other. This means that if the kind is
      "topic" and the passed values are "value A", "value B", this is equal to `topic="value A" or topic="value B"`.
    </ParamField>
  </Expandable>
</ParamField>

<ParamField body="limit" type="integer" required>
  Parameter needed for pagination. It indicates the number of rows per page.
</ParamField>

<ParamField body="offset" type="integer" required>
  Parameter needed for pagination. It indicates the offset to be applied to the first rows returned.
</ParamField>

# Response

<ResponseField name="total" type="integer">
  Needed for pagination. The total number of rows available.
</ResponseField>

<ResponseField name="offset" type="integer">
  Needed for pagination. It indicates the offset to be applied to the first rows returned.
</ResponseField>

<ResponseField name="limit" type="integer">
  Needed for pagination. It indicates the number of rows per page.
</ResponseField>

<ResponseField name="data" type="object[]">
  The data per interaction.

  <Expandable title="properties">
    <ResponseField name="id" type="string">
      The interaction id. It uniquely identifies the interaction and can be used to retrieve further information.
    </ResponseField>

    <ResponseField name="input_text" type="string">
      The input text written by the user.
    </ResponseField>

    <ResponseField name="output_text" type="string">
      The response generated by the assistant.
    </ResponseField>

    <ResponseField name="user_query" type="string">
      A detailed version of the user intent in a single sentence.
    </ResponseField>

    <ResponseField name="end_user" type="string">
      The end user interacting with the LLM.
    </ResponseField>

    <ResponseField name="conversation_id" type="string">
      The conversation id. It uniquely identifies the conversation the interaction belongs to.
      This information can be sent via API in order to group interactions, otherwise it will be generated automatically by Nebuly.
    </ResponseField>

    <ResponseField name="timestamp" type="string">
      The date when the user input was received by the LLM.
    </ResponseField>

    <ResponseField name="end_timestamp" type="string">
      The date when the assistant response was sent back to the user.
    </ResponseField>

    <ResponseField name="warning_message_v2" type="string">
      Description of any warnings or issues detected with the LLM response.
    </ResponseField>

    <ResponseField name="negative_intent_message_v2" type="string">
      If present, it indicates that the user is having a negative experience withing the conversation.
    </ResponseField>

    <ResponseField name="cost" type="float">
      Cost of the interaction, in microdollars.
    </ResponseField>

    <ResponseField name="latency" type="float">
      Time elapsed between the user input and the assistant response, in seconds.
    </ResponseField>

    <ResponseField name="text_language" type="string">
      Language detected in the user input.
    </ResponseField>

    <ResponseField name="verbosity" type="integer">
      Number of characters in the output.
    </ResponseField>

    <ResponseField name="pii_removed" type="boolean">
      Indicates whether PII was removed from the interaction.
    </ResponseField>

    <ResponseField name="feedback_action_text" type="string">
      Text content accompanying feedback actions, if any.
    </ResponseField>

    <ResponseField name="feedback_action_rating" type="number">
      Numerical rating provided as feedback (e.g., 1–5).
    </ResponseField>

    <ResponseField name="feedback_actions" type="string[]">
      Actions taken or suggested via user feedback (e.g., "thumbs\_up", "thumbs\_down").
    </ResponseField>

    <ResponseField name="sentiment" type="string">
      Sentiment classification of the interaction, if available (e.g., positive, neutral, negative).
    </ResponseField>

    <ResponseField name="emotions" type="string[]">
      List of detected emotions from the user input.
    </ResponseField>

    <ResponseField name="tags" type="object">
      Key-value pairs showing custom tags associated with the interaction.
    </ResponseField>

    <ResponseField name="custom_columns" type="object">
      Key-value pairs showing custom columns associated with the interaction defined inside the platform.
    </ResponseField>

    <ResponseField name="custom_aggregates" type="object">
      Key-value pairs of manually defined aggregates associated with the interaction defined inside the platform.
    </ResponseField>

    <ResponseField name="rag_details" type="object">
      Detailed metadata about each RAG item (e.g., source, query, result).
    </ResponseField>
  </Expandable>
</ResponseField>