5+1 Smart Strategies to Track AI Builder Consumption in Microsoft Power Platform

In today’s AI-driven world, Microsoft’s AI Builder empowers Power Platform users to seamlessly incorporate artificial intelligence into apps and automated workflows—without writing complex code. However, AI Builder operates on a credit-based licensing model, meaning that each prediction, classification, or form processing event consumes a portion of your purchased capacity.

If you're a Power Platform admin or Center of Excellence (CoE) lead, tracking this consumption is essential for both cost control and strategic resource planning. In this article, we’ll explore six powerful ways to track and manage AI Builder consumption with real-world examples, ranging from basic reporting to advanced telemetry.

1. Use the Power Platform Admin Center for Overview Reporting

The Power Platform Admin Center is the first place you should go to get an overview of AI Builder usage. Under Resources → Capacity, you’ll find the “AI Builder consumption report,” which shows:

• Total credits used over time (up to 30 days)
• Usage broken down by environment
• Credit consumption by model type (form processing, prediction, etc.)

This tool is ideal for high-level trend monitoring and can help detect unusual spikes in activity.

Example use case: Suppose you see an unexpected jump in credits used last week. By filtering by environment, you discover that a developer accidentally deployed a test model in a production environment, leading to excess usage. This insight lets you address the root cause and optimize deployment practices.

2. Leverage Dataverse Analytics for User and Model-Level Breakdown

The Dataverse tables that underpin AI Builder activity contain far more detail than what’s shown in the Admin Center. One of the most important tables is the AI Event table (msdyn_aievent), which captures:

• What model was used (msdyn_aibuildermodelid)
• How many records were processed (msdyn_inputrecordcount)
• When the event occurred (createdon)
• Who triggered it (ownerid)
• How many credits were used (msdyn_creditsused)

You can view this table in the Power Apps Maker portal by selecting your environment, going to Dataverse → Tables, and searching for AI Event.

Example use case: Let’s say your finance team asks for a breakdown of credit usage by department. By joining the AI Event table with user profile data, you can segment model usage by business unit and generate internal chargebacks or cost reports.

3. Build Custom Dashboards in Power BI for Real-Time Insights

To move from raw logs to actionable intelligence, you can connect Power BI directly to your Dataverse environment and build a live dashboard.

Recommended visuals include:

• Daily/weekly AI credit usage by environment
• Top 5 AI models by consumption
• Heatmaps of model usage by time and user
• Credit trend forecasting

Use the Dataverse connector in Power BI and select tables like msdyn_aievent and msdyn_aibuildermodel.

Example use case: Your governance team wants to monitor whether AI Builder usage aligns with business priorities. Your dashboard shows that an outdated model is still processing hundreds of records per day. That data supports a decision to disable or retrain the model.

4. Automate Usage Reports with Power Automate

With Power Automate, you can create scheduled flows that automatically:

• Query the AI Event table
• Compile a usage summary
• Email a report (PDF, Excel, or HTML) to key stakeholders

You can use the “List Rows” Dataverse connector to pull data and then use Office 365 connectors to email a formatted summary on a schedule.

Example use case: Every Monday, your CoE receives a weekly usage snapshot that shows AI Builder activity broken down by user, credits used, and number of predictions. This automation saves time and ensures consistent governance reporting.

5. Strengthen Governance with Microsoft Purview and Azure Monitor

For organizations with strict compliance and data governance needs, tracking AI Builder consumption is just one part of the puzzle. Microsoft Purview and Azure Monitor can extend your oversight:

• Use Purview to classify data flowing through AI models and enforce usage policies.
• Configure Azure Monitor Logs to track API calls or credit usage telemetry in more complex environments.

These tools are especially useful for organizations in regulated industries like healthcare, finance, or public sector.

Example use case: Your data team uses Purview to label all AI model inputs that contain personal data. If a model tries to process a sensitive dataset without the correct permissions, Purview flags the issue—preventing compliance violations before they happen.

6. Dive into the AI Event Table for Granular Usage Analysis

The AI Event table (msdyn_aievent) is a hidden gem for any organization looking to do deep-dive consumption analysis. Accessible via Power Apps, Power BI, or the Dataverse Web API, it logs every model interaction, making it a perfect foundation for advanced reporting.

Key fields you can analyze:

How to access:

• Go to https://make.powerapps.com
• Select your environment → Dataverse → Tables
• Search for AI Event
• Optionally use Power BI or Advanced Find to filter, sort, and export this data

Example use case: You want to identify which users are using AI Builder the most over the past 90 days. By grouping AI Event records by ownerid and summing msdyn_creditsused, you generate a leaderboard of model usage. This can help with internal training, budget planning, or even setting consumption limits.

Summary

AI Builder is a powerful tool that brings intelligent automation into the hands of business users. But with its credit-based model, tracking consumption is vital—not only to avoid cost surprises, but also to optimize performance, promote responsible AI usage, and ensure compliance.

By combining the built-in admin tools, Dataverse data, custom dashboards, automation flows, and governance solutions like Purview, your organization can implement an enterprise-grade strategy to monitor and manage AI Builder usage effectively.

Source URL: 5+1 Smart Strategies to Track AI Builder Consumption in Microsoft Power Platform