Compute Views by Role

The Compute landing page adapts to how you work with compute resources, offering two persona-based views: Platform Admin View and Data Engineer View. Each view surfaces the metrics and actions most relevant to that role, so you don't have to dig through unrelated data to find what matters to you.

Why this matters

Platform admins and data engineers care about different things when they open Compute — admins need budget and cost-spike visibility across the organization, while engineers need to know which queries and jobs are actually causing problems. Two dedicated views mean each role lands on the information it needs immediately, instead of a single generic dashboard that serves neither well.

Platform Admin View

The Platform Admin View focuses on cost optimization, operational efficiency, and budget management.

With the Platform Admin View, you can:

See cost spikes and long-running jobs as they happen.
Review a detailed cost savings analysis based on ADOC's recommendations.
Check health metrics for overall system performance.
Access chargeback and budgeting information without leaving the page.

The view includes a Copilot natural language interface, letting you ask questions about your compute environment and get real-time suggestions. A Platform Selector lets you scope the entire view to a specific compute platform or environment.

Key panels on this view:

Panel	What It Shows
Estimated Annualized Savings	The potential yearly savings if you implement ADOC's current recommendations. Select it to open a detailed breakdown.
Cost Overview	Your contractual budget for compute resources (Contract Amount), how much of it has been spent (Current Spend), and how much time remains in the current budget period (Time Remaining).
Spend Over Last 7 Days	Daily spend for the past week, broken down by data source.
Budget/Chargeback	A summary of budgeting and cost recovery for your compute usage, linking to the full Chargeback & Budget page.
Recommendations	The latest cost and performance recommendations for your Snowflake and Databricks sources.

Example: If your Current Spend shows you're overspending relative to your Contract Amount, you can drill into Cost Overview to see which data source is driving it, then check Recommendations for specific actions to bring it back in line.

Data Engineer View

The Data Engineer View focuses on job performance, query optimization, and system efficiency.

With the Data Engineer View, you can:

Compare successful versus failed queries over the last 7 days.
Identify expensive and long-running queries.
Review resource optimization opportunities.
See alerts for critical issues as they're raised.

Key panels on this view:

Panel	What It Shows
Successful vs Failed Queries	Query outcomes over the last 7 days. Selecting this panel takes you to Query Studio for a specific data source, where you can investigate further.
Number of Expensive Queries	A count of queries exceeding a cost threshold you define (for example, queries costing more than $1). Selecting this panel opens Query Studio for a data source you choose.
Number of Long Running Queries	A count of queries exceeding a duration threshold you define (for example, 30 minutes). Selecting this panel opens Query Studio for a data source you choose.
Operational Efficiency Score	A single score summarizing the overall health of your compute environment. Select the chart to see the breakdown by data source.
Alerts	Recent critical issues and unexpected data activity, so you can respond quickly.
Recommendations	The latest cost and performance recommendations for your Snowflake and Databricks sources.

Example: If Number of Long Running Queries is higher than expected, select the panel to open Query Studio and identify which specific queries are responsible before they affect downstream jobs.

What's next

Compute Monitors – Set up alerts on the metrics shown in these views.
Compute | Recommendations – Review the recommendations referenced in both views in detail.
Compute | Chargeback & Budget – Manage the budgets referenced in the Platform Admin View.

Last updated on

Was this page helpful?