AI SmartAssist

AI SmartAssist is an in-product chat assistant that helps you write queries, work with data, create visualizations, and troubleshoot issues directly from SquaredUp.

Wherever a code editor appears in the app, SmartAssist can analyse the contents of your current session and provide suggestions intelligently tailored to what you are working on. This allows you to move from raw data to meaningful dashboards faster.

SmartAssist reduces trial and error when building dashboards by helping you generate queries more quickly and understand how to shape your data for visualization.

Enabling SmartAssist

To access SmartAssist you'll first need an administrator to enable it for your organization by doing the following:

1. Navigate to the Settings > Organization page.
2. Enable the Enable AI features toggle.
3. Click Save.
   

Accessing SmartAssist

You can open the SmartAssist chat panel wherever a code editor appears in SquaredUp by clicking Open SmartAssist

• The Parameters tab of the Tile Editor when working with data streams that require a query / request.
• The Query tab of SQL Analytics.
• When creating scripts via the Settings > Scripts > Add script window.
  

Using SmartAssist

You can ask SmartAssist questions using simple prompts written in plain language. SmartAssist analyses the contents of the editor before responding, so its suggestions are tailored to the query, script, or expression you are currently working on.

This allows you to focus on the result you want rather than the exact syntax needed to achieve it.

SmartAssist sessions remain active for the duration that you are editing a tile, meaning that you can freely close and open the chat panel and the conversation will persist. However, exiting the tile editor at any point will reset the session and clear the chat history.

Building and refining queries

SmartAssist helps you move through each stage of configuring a tile by generating queries, improving filters, explaining logic, and reshaping results as you go.

Examples:

• Show incidents created in the last 7 days grouped by priority
• Filter this to only include Sev1 and Sev2 alerts
• Return daily totals instead of hourly
• Add a column showing average resolution time per service
• What does this WHERE clause do?

These kinds of prompts help you quickly move from selecting a data stream, to refining parameters, aggregating results, and shaping data in SQL Analytics, without needing to write or interpret complex query logic manually.

This makes it easier to adjust queries confidently, especially when working with unfamiliar datasets or query languages.

Code suggestions

When working with scripts or queries, SmartAssist often provides queries or scripts for you to use, or suggests improvements to your existing code.

To apply suggestions:
Click Paste to apply the suggestion and replace your own content with the SmartAssist code snippet.

To undo a suggestion:
Press CTRL + Z to undo the a suggestion you applied and revert to your original content.

To re-do suggestions:
You can reapply any previous suggestions from the chat window by simply clicking Paste again.

SmartAssist diff pane

When presented with a code snippet suggestion, click Preview to open the suggested changes in Preview Changes window.

From here you can view both your content and the suggestion side by side, or toggle to an inline view by clicking the Show inline button.

If you are happy with the changes, click Apply to confirm them and overwrite your content, otherwise click Cancel to exit the window.

Troubleshoot SQL issues

SmartAssist can help identify why SQL queries fail or return unexpected results by analyzing the current editor content, suggesting likely causes and the recommend changes to correct them.

Examples:

• Why is this query failing?
• Why am I getting duplicate rows?
• Why am I not getting results for this timeframe?

Preparing data for visualization

If you find that the data you have doesn't quite work for the visualization you're trying to achieve, SmartAssist can suggest how to reshape results so they fit. This helps expand the capabilities of your dataset and help you find insights you may otherwise have missed.

Examples:

• Update this query for a time series visualization
• Group this data for a bar chart
• Sum the first column for a Scalar visualization

Tips for getting the best responses

While SmartAssist is intelligent enough to respond to simple prompts using immediate context, there are a few things you can do to further improve the responses you get:

Be specific about what you want

Laying out the clear goals of what you are trying to achieve greatly helps SmartAssist generate more accurate suggestions.

Instead of:
Fix this query

Try:
Fix this query so it returns incidents created in the last 24 hours

If you already know how you want the data to be structured, include those details in your request.This helps SmartAssist generate queries that are closer to your intended visualization:
Group results by service and return daily totals for the last 7 days

Add follow-up prompts

SmartAssist works conversationally, so you can continue improving responses step by step:

• Now filter this to production only
• Change this to show weekly totals instead
• Only include closed incidents
• Rename this column to Resolution time

This is often the quickest way to get to the result you want. Rather than starting over with a new prompt each time, you can build on the previous response and gradually shape the query, expression, or output until it matches your needs.

Was this article helpful?