Automation Anywhere Agentic Process Automation Reviews & Product Details

What I like best about Automation Anywhere’s Agentic Process Automation is how it combines AI agents, RPA, and orchestration to automate complete business workflows. Its ability to handle unstructured data and adapt to exceptions makes it useful for processes that require more than fixed rules. I also appreciate the human oversight and governance features, which allow teams to retain control over critical decisions. Together, these capabilities can reduce manual effort and help employees focus on higher-value work. Review collected by and hosted on G2.com.

I would like a more intuitive interface for configuring AI agents, with clearer guidance on workflow design, exception handling, and when to use agents versus rule-based automation. For integrations, more practical examples and clearer error messages would help reduce setup effort and make connection failures easier to resolve. More detailed performance dashboards showing processing times, failed steps, and resource usage would make it easier to identify bottlenecks and maintain reliable workflows at scale. Clearer licensing and cost estimates, alongside better visibility into AI decisions and human approval steps, would help teams plan budgets and manage oversight more effectively. A more structured onboarding journey, with role-specific tutorials, guided sample projects, and clear support escalation timelines, would help new users become productive faster and resolve issues with less disruption. Review collected by and hosted on G2.com.

The biggest difference in our day-to-day operations is how smoothly the platform handles fuzzy, exception-heavy work without causing downstream jobs to crash. The visual builder makes it very intuitive to set up prompt templates and human review checkpoints, and the native connectors to our core databases and enterprise systems eliminated weeks of integration headaches. Instead of writing brittle rules for every small format shift, the AI agents evaluate unstructured input on the fly and reliably trigger the right bot actions.

Onboarding was faster than we expected because the pre-built templates gave our developers a real running start, and support was quick to jump in on configuration questions during rollout. Even though the platform licensing is a significant budget commitment, the ROI became clear once our team stopped spending hours each week manually sorting through edge cases, while still keeping compliance satisfied with clear audit trails. Review collected by and hosted on G2.com.

The biggest friction point for me is debugging and overall visibility when an agent behaves unexpectedly. With classic rule-based automation, a failure usually points straight to a broken line, a timeout, or another clear culprit. But once AI reasoning is driving decisions, understanding why an agent misread an edge case or chose the wrong path often turns into tedious trial and error. As a result, troubleshooting feels much heavier and less deterministic.

On top of that, the licensing structure, combined with AI consumption costs, makes total spend difficult to predict as workloads scale. And while it’s marketed as low-code, it still requires solid engineering oversight to tune prompts and maintain fallback logic, so non-technical teams can’t realistically manage it on their own. Review collected by and hosted on G2.com.

What I like best is the consolidated platform. RPA, document AI, process discovery, human-in-the-loop forms, analytics, and now AI agents all sit in one Control Room, and it is easy to use for everyone: business users, developers, and IT managers. I have worked with Automation Anywhere for 9+ years across banking, government, healthcare, and oil and gas, and this is the version where the pieces finally feel like one product. Review collected by and hosted on G2.com.

Local LLM options: For clients who cannot send data outside the country, connecting agents to self-hosted or in-region models works, but it takes more setup than it should. More out-of-the-box support for regional and self-hosted model hosting would be very welcome.

None of these stop us from recommending the platform, and the product team has been responsive when we raise feedback through the partner and MVP channels. Review collected by and hosted on G2.com.

Automation Anywhere Agentic Process Automation is the software aimed at improving processes where a series of repetitive actions is spread across various apps. I use it for automating actions in processes, orchestrating actions within different systems and minimizing manual effort in repeating business operations. As a result, process execution becomes more consistent while people have extra time to work on the processes that need human intelligence. Review collected by and hosted on G2.com.

One problem occurs in case a process is characterized by many exceptions or has changing business rules. In such case, it requires proper configuration and testing since an automated process needs to know how to behave if the situation deviates from the standard scenario. Review collected by and hosted on G2.com.

I like that Automation Anywhere Agentic Process Automation combines AI agents, generative AI, and traditional RPA to automate complex workflows end to end. It works well with unstructured data and helps extract useful information from documents and text. I appreciate being able to integrate agents with enterprise applications like ServiceNow and orchestrate work across multiple applications, with less manual intervention and fewer repetitive tasks. The AI agent orchestration helps break complex processes into smaller tasks and coordinate them. I also value human-in-the-loop validation for critical decisions, along with monitoring and governance that help me maintain control over automated processes. These capabilities help me build scalable, reliable automation workflows and improve productivity. Review collected by and hosted on G2.com.

I think developing and debugging complex agentic workflows could be simpler, especially when multiple AI agents and enterprise systems are involved. I’d like better debugging capacity, error handling, retry mechanisms, visibility, and troubleshooting. Optimizing LLM response latency and token usage could also help reduce operating costs. I’d also find the platform more useful with more flexible third-party integrations and better support for custom LLMs. Review collected by and hosted on G2.com.

What I find most promising about Automation Anywhere Agentic Process Automation is its approach to combining AI-driven decision-making with traditional automation. It has the potential to handle multi-step workflows, coordinate tasks across different systems, and reduce repetitive manual effort. I also like the broader focus on moving beyond rule-based automation toward workflows that can interpret context and adapt to different scenarios. From a business perspective, this can help improve productivity and allow teams to focus on more strategic work, provided the workflows have appropriate monitoring, guardrails, and human oversight. Review collected by and hosted on G2.com.

One area that could be improved is the complexity of designing, testing, and maintaining agentic workflows, particularly when processes involve multiple systems or unexpected exceptions. AI-driven automation also needs strong guardrails, clear decision traceability, and reliable human-in-the-loop controls to ensure consistent outcomes. Troubleshooting failures and understanding why an agent took a particular action should be straightforward for users. Better onboarding, simpler debugging tools, and more transparent monitoring would make the platform easier to adopt and manage at scale, especially for teams moving from traditional rule-based automation. Review collected by and hosted on G2.com.

What I like best is how it combines automation with AI-based decision making, so the process doesn't have to be completely rule-based. The platform makes it easier to automate repetitive tasks and handle workflows that need some level of judgement. I also like that the workflows can be customized based on the business process, and the overall setup is fairly straightforward once you get familiar with the platform. Review collected by and hosted on G2.com.

The main downside is that some of the more advanced features take time to understand and configure. When a workflow becomes more complex, it can be difficult to figure out why a particular step is not behaving as expected. The learning curve could be a little easier for new users, especially around debugging and managing more complicated automations. Review collected by and hosted on G2.com.

I think Automation Anywhere APA is a great tool for integrating AI into our processes because it’s very intuitive. You can interact with AA elements like a user or a developer, which makes it easy to work with in different ways. Review collected by and hosted on G2.com.

There are very few elements I don’t like about Automation Anywhere Agentic Process Automation. One of them is that the graphic interface for Forms is quite basic from a visual perspective. It should offer more flexibility for customization and adaptation. Review collected by and hosted on G2.com.

What I like best about Automation Anywhere Agentic Process Automation is its ability to combine AI agents with automation to handle complex, repetitive business processes with minimal human intervention. Its integration of generative AI, intelligent decision-making, and workflow automation helps improve efficiency, reduce manual effort, and streamline end-to-end business operations. I also appreciate its potential to automate dynamic workflows and support faster, more accurate decision-making. Review collected by and hosted on G2.com.

One area for improvement is the complexity of configuring and managing agentic workflows, especially when integrating with existing enterprise systems. Debugging AI-driven processes and ensuring consistent, predictable results can also be challenging. Better documentation, simpler configuration, and more transparent monitoring tools would make the platform easier to use and maintain. Review collected by and hosted on G2.com.

I’ve found Automation Anywhere Agentic Process Automation to be very effective for extracting information such as sentiment, classification types, and tone from unstructured datasets. It helps identify high-priority tickets based on the tone and behavior reflected in users’ requests, which is especially beneficial when it’s integrated into customer service processes. In addition, the AI skill integration has been extremely helpful for pulling the required details from candidate resumes for HR onboarding applications, and similar use cases. Support from AA has also been very good, both in terms of performance and in helping us achieve ROI. Review collected by and hosted on G2.com.

If it offered a way to import input data directly from documents like Excel, PDFs, and DOCX, it could feel like something truly special for Automation Anywhere Agentic Process Automation. Right now, the setup feels fairly basic and has a plug-in style approach, but you still need a solid understanding of the product’s usability to get the most out of it, especially when integrating the product with others. Review collected by and hosted on G2.com.