Agentic Process Automation System (Now Certified for WorkSpaces)

I have been using Automation Anywhere for about two years, mainly for automating repetitive workflows and exploring how RPA can streamline routing business processes.

My main use case for Automation Anywhere involves automating repetitive optional tasks. For example, I work on a process where data comes in through Excel files and normally someone had to manually validate it and enter it on internal systems and generate a report. With Automation Anywhere, we automate most of the workflows by reading the Excel data, validating the required fields, updating the systems, and generating the report fields. It reduces a lot of repetitive manual work and also helps us to avoid basic data entry errors.

Apart from the Excel data entry workflow, I use Automation Anywhere for report generation, moving data between different systems, and automating repetitive validation tasks. One thing I found useful is that you can combine multiple small manual steps into a single end-to-end workflow. Instead of someone consistently switching between Excel, internal applications, and a report, the bot can handle most of the process and only involves a person when something actually needs attention.

The best features Automation Anywhere offers for me include a visual workflow builder, reusable bots and components, and the way it integrates with common business tools including Excel, web applications, and internal systems. I also appreciate the central Control Room because it makes it easier to manage bots, monitor execution, and handle failures. Another useful part is data, document, and AI-related automations, especially when a process involves extraction and validation of data from semi-structured documents.

The flexibility of Automation Anywhere's truly universal orchestration is one of its strongest points. We are not limited to automating with just one application; we can orchestrate workflows across web applications, APIs, databases, documents, and other enterprise systems. I also appreciate that the orchestration layer does not force everything to depend on AI models. Different models or AI services can be incorporated depending on the use case, while Automation Anywhere manages the broader workflow. In our experience, this is useful for combining traditional rule-based automation with AI-driven steps while still keeping monitoring, validation, and human approval in the same end-to-end processes.

One area that Automation Anywhere can improve is the debugging experience. When you have a long workflow with multiple bots and integrations, finding the exact cause of a failure can sometimes take longer than expected because step-by-step debugging and clear errors would help. I would also appreciate seeing some of the more advanced features become easier to configure. The platform is powerful, but there can be a learning curve when you move from basic automation to complex workflows. Improving that would make development and maintainability faster.

I think visual management and collaborative work can also be smoother, especially when multiple developers are working on the same automation for a longer project. Comparing changes and managing different versions can become a little cumbersome. I would say improving the overall performance of the development environment for bigger workflows as automation becomes more complex, keeping the interface fast, and making dependencies easier to manage would improve the development experience quite a bit.

I have been working with Automation Anywhere for two years.

Automation Anywhere is pretty stable in my perspective because we do not see many problems. Most problems we encounter arise from changes in external applications, UI elements, credentials, or input data rather than Automation Anywhere itself.

The scalability of Automation Anywhere has been good for us. As the number of workflows and bots executing increases, we are able to manage them centrally through the Control Room without needing to redesign the whole setup. The cloud-based model also makes it easy to add more bots, capability, and capacity as demand grows. Our considerations around experiences and maintainability usually center on licensing, infrastructure planning, and ensuring workflows are designed efficiently rather than being limited by the platform itself.

The customer support from Automation Anywhere has been very good overall. Whenever we need help with configuration or troubleshooting, the team has been responsive and generally provides useful guidance. For more complex technical issues, resolution sometimes takes longer because there are investigations involved, but overall, I rate the support experience positively at eight out of ten.

We did not use any full RPA platform before Automation Anywhere. Most processes were either handled manually or through small scripts and application-specific automations.

My experience with Automation Anywhere's pricing, setup cost, and licensing is pretty positive. The pricing is more on the enterprise side, so I would not consider it a low-cost solution, especially once you scale to many bots and users. However, for a process with enough volume, the time saving and reduction in manual effort can justify the cost. Initial setup took some planning, particularly around bot configuration and integration, but after that, it becomes fairly manageable. Licensing can be a little complex because the cost depends on the capabilities and scale you need.

We have seen a positive return on investment with Automation Anywhere mainly through the time saving and reducing manual efforts. For some repetitive workflows, tasks that previously required one or more hours of manual work can now be completed much faster and with minimal human involvement. Document processes have reduced the time spent on extracting, validating, and entering data. I do not have an exact professional financial ROI percentage because I am not directly involved in the financial reporting, but from an operational perspective, the reduction in manual hours and errors has been worthwhile for us.

My experience with Automation Anywhere's pricing, setup cost, and licensing is pretty positive. The pricing is more on the enterprise side, so I would not consider it a low-cost solution, especially once you scale to many bots and users. However, for a process with enough volume, the time saving and reduction in manual effort can justify the cost.

We were looking at a few alternatives before choosing Automation Anywhere, mainly UiPath and Microsoft Power Automate. We compared them around ease of development, integration capabilities, centralized management and scale, and overall enterprise governance. Automation Anywhere stood out for us because of its cloud-based approach, central Control Room, and capability to manage large end-to-end automation workflows; those other solutions were not able to do it.

My impression of Automation Anywhere's AI governance features is positive. They give us better control over how AI is used within our automation, especially around access to sensitive data and traceability. Role-based permissions and audit trails are important because we can control who is building, modifying, or running particular automation and have visibility into what happened. For data integration, we use validation rules and human review for sensitive and low-confidence cases, rather than relying completely on AI output. I see the governance features as a strong foundation, but they work best when combined with the organization's own compliance and data handling policies.

The biggest challenges with agentic AI for me are governance and security, reliability, and maintaining control over what an AI agent is allowed to do. Unlike traditional rule-based automation, an AI agent can make dynamic decisions, and we need to be confident about its output, data access, and what happens when it comes to making a mistake. Automation Anywhere helps by combining AI capabilities with more controlled automation. Workflows can handle access control, maintain auditability, and include validation rules and human approval for sensitive steps. I think this combination is important because we want the flexibility of agentic AI without sacrificing governance and predictability that we already maintained with traditional automation.

The main objective with agentic process automation was to move beyond automating only fixed repetitive tasks. We want the automation to handle situations where there are expected changes in input or some level of decision-making involved. For example, in document processing, instead of just extracting predefined fields, the goal is to understand the outcome of the information, validate it, decide what should happen next, and route exceptions to the right person. Agentic process automation helps us connect these fields into more intelligent end-to-end workflows while keeping human oversight for critical decisions.

My impression of the data advantage provided by Automation Anywhere's process reasoning engine is that it really excels because it is designed around enterprise workflows rather than just general-purpose AI tasks. This gives us a more useful understanding of how different steps, applications, data, and business rules fit together in a process. Whenever I use it, the most valuable aspect is in reducing the amount of context we have to provide manually and helping the system make better decisions around the next steps in the workflow. I still validate a decision for critical processes, but overall, enterprise process context makes agentic automation particularly relevant for real business use cases.

My advice would be to start with processes that are repetitive, high-volume, and have clear rules rather than trying to automate everything at once. Pick a few workflows where you can clearly measure the time saving and then expand from there. I would also recommend putting proper exception handling, logging, governance, and monitoring in place from the beginning. For AI-based automation, keep human validation for sensitive, low-confidence decisions. Automation Anywhere is very capable, but the quality of the results still depends heavily on how well you design and govern the automation. I gave this review an overall rating of nine out of ten.