Over the last few years, Power Platform has increasingly been filled with AI components designed to assist various business cases. AI features make work quicker and add value to end products by using unpredictable logic that classic logic could not apply at all or would require significant effort to apply.

In-App Agents are assistants that help perform manual tasks in model-driven applications built on the Dataverse database. They are very easy and fast to enable for use in the system, as they serve as additional modules.

It filters data in a table based on natural language and displays data based on the chosen criteria.

For example, the librarians work with model-driven systems and constantly search for books to track their availability. Every day, visitors come and ask for books they know some details about but do not know the title of. They spend a lot of time finding the book by applying standard filters. Although we include the Data Exploration Agent in the librarian app, it will make the searching process for unique requests faster.

It quickly visualizes table data as a chart and provides a report in the desired data and format (bar chart, pie chart, etc.).

If there is a sales company that wants to evaluate the sales results, but they do not want to maintain additional BI licenses for something they barely use. They can include the Data Visualization Agent, as it is quick to implement and requires no additional licenses.

It analyzes the records the user previously created and proposes the most likely values to save time when creating records manually.

In case, PM often repeats itself in the internal project management system, creating the same tasks over and over again. They use Data Entry Agent, which analyzes the tasks PM usually creates and proposes the most probable values. The PM spends less time on creating tasks if the fields are prefilled.

It creates a text or document report from the row. The agent can summarize all the information regarding the record, and then app users can review the main points relevant to them and even receive a generated document report with the required structure.

Include the Row Summary Agent to generate text or document summarization based on sales records. Managers, when they look at the table of records with a big budget, can click on the needed record to generate a report that lists, e.g., how many days the maximum norm is to deliver the order and how much of this order is being delivered. It will speed up the work, as the user will have all relevant information in one place.

This is a new feature of vibe coding. AI in Power Platform can generate a custom page for a model-driven application based on the request: what can the user do and how shall the page look?

This is a way to implement the application’s new functionality much more quickly than through manual development. As a custom page, the page can connect to the business database to work with data. The only concern is that the page’s interface may vary slightly despite the request.

An app maker has added a data source connection to Dataverse and written a prompt of what should exist. The practical page appeared in a matter of seconds. If an app maker wants to change something, they can chat with AI to request additional elements, a color palette, etc. If an app maker is a skilled developer, they can modify the source code.

Generative Page is not the only fully generated AI component that exists. Microsoft also proposes vibe coding with Code Apps. If Generative Page replaces the custom page, Code Apps replaces the standard web app. It is a back-end development that can create a flexible application that corresponds to business compliance rules.

In-app agents are AI-powered bots designed to streamline record management within model-driven apps built on Dataverse, helping users navigate the interface, generate quick reports, and access real-time information effortlessly. This efficiency is further enhanced by Generative AI Pages, which allow for the rapid creation and implementation of custom pages, effectively replacing traditional, time-consuming development processes. Additionally, the ecosystem includes Code Apps, which are AI-generated web applications tailored to meet specific business needs with minimal manual intervention.