Agentic AI RPA

Overview

Appian's Agentic AI RPA brings a new level of intelligence to your processes by combining robotic process automation (RPA) with AI. With AI, building a robotic task in Appian RPA is easier than ever. Instead of manually defining each step, Agentic AI RPA adapts in real time, determines the best next action, and saves you time and effort.

Why use Agentic AI RPA?

Agentic AI RPA makes automation smarter and more intuitive. By combining artificial intelligence with robotic process automation, it overcomes technical barriers and simplifies automation across different software systems using smart screen vision. Smart screen vision is the technology that allows robotic tasks to recognize and interact with on-screen elements, making it possible to process visual data just like a human would.

Many automation processes are designed to handle routine or predictable tasks, with workflows that align to specific business needs. When those needs evolve, workflows must be updated, requiring additional development time. Agentic AI RPA reduces this effort by making dynamic adjustments on its own.

Key benefits

• Faster automation setup: Reduces reliance on lengthy robotic tasks and rule-based configurations.
• Adaptive decision-making: Adjusts workflows in real time based on changing data.
• Automate any software system: Works with legacy systems like Mainframe and Citrix.
• Lower development effort: Automates complex processes without requiring deep technical expertise so you can spend less time on setup and more time on other important tasks.
• Seamless RPA integration: Works with Appian RPA agents to execute actions efficiently.

Example use cases

Stay resilient to unexpected UI changes

Robotic tasks are designed to follow expected patterns, but when a user interface changes, they may no longer function correctly. Traditionally, developers would need to manually update the robotic task to account for these changes—like a redesigned login page or a button moving to a new location. Agentic AI RPA eliminates this overhead by automatically detecting and adapting to interface changes, allowing robotic tasks to continue running smoothly without requiring frequent updates.

This means fewer manual fixes, lower maintenance costs, and uninterrupted automation, even when business-critical systems change.

Automate any Windows software—even legacy systems

Automating Windows software systems like SAP, Mainframe, MS Office, and Citrix-based apps can be challenging. Agentic AI RPA solves this by combining smart screen vision with AI-driven decision-making, allowing you to automate tasks—even when field IDs are missing or unpredictable.

For example, automating an SAP transaction used to require complex workarounds like keyboard shortcuts. Now, Agentic AI RPA can navigate the interface, fill in fields, and complete transactions reliably, reducing manual intervention, streamlining development, and improving automation resilience.

Handle business exceptions dynamically without complex workarounds

Managing business exceptions in automation can be tricky. Traditional approaches rely on workarounds, making development more complex, maintenance harder, and automation less resilient. Agentic AI RPA simplifies exception handling with intelligent, dynamic responses.

For example, if a robotic task tries to access something that isn't available or has changed, Agentic AI RPA can recognize the issue and keep the robotic task running instead of stopping.

This approach reduces the need for manual error handling, making automation more robust, easier to maintain, and adaptable to real-world conditions.

Connecting with other actions

Agentic AI RPA works with existing low-code actions to enhance automation. Here are a few examples:

• Login Automation: Use Agentic AI RPA to log in to a system, then continue the workflow with other actions to navigate or extract data.
• Data Extraction: Capture details from a desktop app and store them in a variable for use in an Appian process model.
• Error Handling: If an action encounters an issue (e.g., a missing field), the process can continue by dynamically adjusting to the interface changes.

What is the Agentic AI RPA Beta Program?

The beta program gives selected customers and partners hands-on access to Agentic AI RPA in a non-production environment. The program runs through August 2025.

Who should join the beta

Ideal use cases for this feature are similar to existing RPA scenarios, such as:

• Managing workflows in healthcare
• Processing documents in banking

Agentic AI RPA helps participants achieve results faster and extends automation to systems that are typically harder to integrate.

Benefits for participants

• Early access to Appian's latest features
• Direct feedback sessions with Appian Product and Engineering
• Early learning opportunities to reduce future implementation time
• Free support from Appian experts on use cases and best practices

Cost

No additional licenses are needed for the beta. Advanced Tier licenses are required for production use after general availability.

Who needs to be involved

At least one Appian developer is required. It's also helpful to involve product owners or business users to assess feature fit, and a senior leader to support adoption after general availability.

Time commitment

Time depends on use case complexity and developer availability. With just a few hours a week, a developer can demonstrate value within a few weeks.

Providing feedback

Participants meet regularly with the Appian product team (30 minutes weekly or 60 minutes every other week) to share feedback and ask questions. Participants should document feedback to discuss during these meetings.

Legal agreement

All participants must sign a beta agreement, separate from any existing contracts. If your organization signed a beta agreement within the past year for another program, you don't need to sign it again.

What happens after the beta

Beta findings will be shared on the Appian Community beta page.

As a beta participant, you can continue using the feature in one of your non-production environments at no extra cost after it becomes generally available. Note that the sites provided during the beta program will close when the program ends.

How beta access is provided

Appian will provide an Appian-owned cloud site located in the US-West-2 (Oregon) region for beta access.

Beta updates

Beta functionality will be updated as needed to fix critical bugs and introduce new capabilities during the beta period.

Technical requirements

There are no technical requirements during the beta, since Appian provides the site. After general availability, you'll need:

• A cloud site (not self-managed) based in a US region or with cross-region support
• The latest version of the Appian low-code platform and RPA

Use the Agentic AI RPA low-code action

The RUN AGENTIC AI TASK action allows you to set up a robotic task using natural language. This action appears in the RPA actions palette and integrates seamlessly with existing RPA actions.

Add the Agentic AI RPA action to a robotic task

1. Complete one of the following options:

   If you want to…	Then…
   create a new robotic task	complete the steps for creating a robotic task then continue to step 2.
   add to an existing robotic task	in the list of design objects, find the robotic task you want to configure in the list and click its name. A new window displays and you are brought to the robotic task configuration screen.

2. Drag and drop the RUN AGENTIC AI TASK action from the RPA actions palette onto the Action Configuration pane.
3. In the CONFIGURATION area of the pane, configure the following parameters:
   • Input Variable (optional): This variable passes data to the AI model, allowing it to adjust dynamically. Use this field to define a key parameter AI should consider when executing the robotic task. For example, you could use documentPath to locate specific files for processing. You may use one input variable per action.
   • Input Credentials (optional): Use this field to securely provide login credentials or authentication details. This ensures the robotic task can access protected systems or programs. You may use one input credential per action.
   • Prompt (required): Enter a clear, natural language instruction to define the goal of this action. For example, "Log in to SAP using these credentials", "Extract customer data from this form", or "Check if an invoice is approved and store the result."
   • System capabilities: This parameter provides background instructions or context to guide how the robotic task executes the action. For example, "Always use secure authentication methods and retry once if login fails" or "Ensure all required fields are completed before submitting".
4. In the OUTPUT area of the pane, configure the following parameters:
   • Operator: Select is stored as or is appended to.
   • Target: Select the robotic task variable where you want to store the result. If the action extracts information, it automatically saves the data in a variable for later use.

5. In the AFTER COMPLETION area of the action pane, configure the following parameters:

   • Wait before executing next action (seconds): Enter the number of seconds the robotic task waits before it continues to the next action in the robotic task definition.
   • Take a screenshot: If you select this option, the robotic task takes a screenshot of the host machine after the action completes. If the action failed, the screenshot is taken immediately before the exception is thrown and cleanup is started. If the action completed successfully, the screenshot is taken immediately after the action is performed.
6. Click SAVE CHANGES in the page toolbar.

Install an RPA agent

To run Agentic AI RPA robotic actions, you need to have an RPA agent installed and running on your host machine. The RPA agent allows your host machine to connect to Appian RPA and execute robotic tasks.

To install an RPA agent:

1. Install the RPA agent on the host machine that will run the Agentic AI RPA robotic tasks.
2. Make sure the RPA agent is running before starting any robotic tasks.

Once the agent is running, Agentic AI RPA can perform tasks on your host machine—detecting screen elements, executing steps, and handling visual data automatically.

Need help installing the agent? See Agents for step-by-step instructions.

Feedback