Embedded conversation starters

Latest Dynatrace
Reference
1-min read

Dynatrace applications like Kubernetes (new) Kubernetes, Threats & Exploits Threats & Exploits, Databases Databases, and Problems app - new Problems allow you to trigger a predefined, contextual Dynatrace Assist prompt to increase your productivity and conversation efficiency.

Prerequisites

To access the application integrations, ensure the following:

Dynatrace Intelligence generative AI has been enabled on the environment level. For details, see Enable Dynatrace Intelligence generative AI on your environment.
You have permissions to access the conversational recommender skill. For details, see User permissions.

Dynatrace Assist in Kubernetes

You can quickly get an explanation of any warning signals with Generative AI in Kubernetes, powered by Dynatrace AI Dynatrace Assist. This allows you to get insights into the event details, typical root causes, and common remediation steps without accessing the documentation or other Dynatrace-related sources directly.

To access this functionality:

Navigate to any list page in the Kubernetes app (such as Clusters, Nodes, Namespaces, or Workloads).
Select any warning signal, and then select Explain warning signal.
Dynatrace Assist will open and auto-execute the predefined prompt.
Generative AI will provide a response that details:
- A general explanation about the event.
- Typical root causes for the event, starting with the most common ones.
- Common remediation steps for each of the root causes.

Dynatrace Assist in Threats & Exploits

Dynatrace Intelligence generative AI can provide contextual, plain-language explanations of detection findings to accelerate understanding and response.

To access the functionality:

In Threats & Exploits, select a finding.
Select Explain finding in the upper-right corner of the finding details pane.

When selected, Generative AI analyzes the technical details of a detection finding and generates a structured summary that may include:

What the finding means: Interprets technical terms and describes the nature of the detected behavior (for example, a process modification, SQL injection attempt, or unexpected kernel code change).
Why it matters: Highlights severity levels (such as CRITICAL) and potential implications for application performance, system stability, or data security.
What to investigate: Suggests next steps such as reviewing affected components, analyzing logs and metrics, and assessing operational impact.
How to respond: Recommends remediation actions and links to relevant tools.

The structure and depth of Generative AI's explanation may vary depending on the nature of the detection and available context. While Generative AI aims to provide detailed insights, not all findings will include every element listed above.

Generative AI explanations are tailored to the nature of each detection—whether it's a code-level exploit, behavioral anomaly, or infrastructure-level threat—providing relevant, actionable insights that accelerate triage and support informed decision-making, even for users without deep security expertise.

Dynatrace Assist in Databases

In Databases , Dynatrace Intelligence generative AI can provide natural language explanations of execution plans, breakdowns of relevant details, and recommendations on how to improve statement performance.

Query execution plans provide detailed information on how a database will execute an SQL query. While these provide the raw data on how to improve query performance and reduce resource consumption, they require expert knowledge to read and interpret. With the Generative AI integration, non-expert database users, such as developers, gain the knowledge they need to optimize their application performance and database utilization.

To summarize an execution plan with Generative AI:

In Databases , go to Explorer.
In the rightmost column, select the statement performance icon.
Expand the statement you would like to improve. If an execution plan is not already available, you can request one.
Select the Execution plan tab, and select Summarize execution plan.
Dynatrace Assist will open and auto-execute the predefined prompt.
Generative AI will provide a response with insight on the selected database execution plan.

Dynatrace Assist in Problems

In Problems Problems app - new , Dynatrace Intelligence generative AI can provide clear summaries of problems, their root causes, and the suggested remediation steps. Generative AI explains individual issues in clear language from the problem details page and can perform a comparative analysis when multiple problems are selected from the list view. This helps you identify common root causes and propose corrective steps without relying on a team of experts or waiting for critical insights.

To explain a single problem with Generative AI

Navigate to any problem detail page.
Select Explain in the upper-right corner of the page.
Dynatrace Assist will open and auto-execute the predefined prompt.
Generative AI will provide a response that details:
- An explanation of what happened.
- Why the problem occurred.
- Actionable steps to remediate the problem.

To explain multiple problems with Generative AI

Navigate to the problem list page.
Select up to 5 problems.
Select Explain above the table.
Dynatrace Assist will open and auto-execute the predefined prompt.
Generative AI will provide a response that details:
- An explanation of each problem and why it occurred
- Actionable steps to remediate the problem
- Any relationship between the problems

Dynatrace Assist in Dashboards

You can use the power of DQL to integrate Dynatrace AI Dynatrace Assist into your Dashboards tiles.

By adding | fieldsAdd prompt and | fieldsAdd execute commands, you can predefine and auto-execute prompts in Dynatrace AI Dynatrace Assist, allowing you to quickly get an explanation about the query results, or receive suggestions on how to improve the query or resolve a problem.

You can also provide additional information to Dynatrace AI Dynatrace Assist via the supplementary context by adding the following:

| parse "{\"result\":[{\"type\":\"supplementary\", \"value\":\"The character`*` often represents sensitive data that has been masked\"}]}", "LD JSON_ARRAY:contexts"
// or for a dynamic context

While the supplementary context is hidden in the chat UI, it can help Generative AI provide better answers for your use case. For example, you can ask Dynatrace AI Dynatrace Assist to use information from a certain field when answering your prompt:

| fieldsAdd supplementaryContext = concat("{\"result\":[{\"type\":\"supplementary\", \"value\":\"Use the following info to answer the question: ", record.summary, "\"}]}")
| parse supplementaryContext , "LD JSON_ARRAY:contexts"

To integrate Dynatrace AI Dynatrace Assist into your Dashboards tiles

Go to Dashboards and open a dashboard you can edit.
Select a dashboard tile that contains a DQL query.
Select Edit to open the edit menu on the right.
In the DQL section of the edit menu, add the following to your standard query:
```
| fieldsAdd prompt = concat("{your question}",  your.field.name)
| fieldsAdd execute = true
```
If you want to predefine the prompt without automatically executing it, remove | fieldsAdd execute = true.
This integration doesn't work for queries with the makeTimeseries command.

To open the integrated Dynatrace AI Dynatrace Assist

Select the next to your chosen field entry.
Select Open with… > Ask a question.

If you've added | fieldsAdd execute = true to your query, the predefined prompt will be executed once you open Dynatrace AI Dynatrace Assist. Otherwise, you'll be able to change or edit the prompt in the message window before manually executing it.

Dynatrace Assist in Notebooks

You can use the power of DQL to integrate Dynatrace AI Dynatrace Assist into your Notebooks.