My main use case for Automation Anywhere is IT automation, specifically IT operation automation.
Agentic Process Automation helps me achieve these goals. The AI Studio helps us create agents quickly. Once I identify a problem, I can create a workflow around it. This may take many days, but this solution helps us quickly build agents, deploy them, and test them with synthetic data, and sometimes with real data as well, which helps us fix problems faster than before in the AI world.
The experience with the AI Agent Studio and its integration feature is important. If the agents are not able to connect to various systems, they will not be able to function efficiently. All you get from other solutions is chat and suggestions on what needs to be done. With Agentic capability, you can actually do those things. It has memory, it has access to tools, and it can do the job on our behalf as opposed to just telling us what to do. Getting things done versus telling us what to do is the difference.
To improve Automation Anywhere, I think we need more integrations with more external tools. Continued support with better, more capable models, more optimal token usage, and reduced hallucinations are some things I want to see.
I have been using Automation Anywhere for a couple of years, three to four years.
I did not notice any significant downtime or crashes. Most of these processes are internal to us, so even with some minimal downtime, it is somewhat acceptable.
We do have a point of contact regarding the Automation Anywhere Center of Excellence. There is a person who has been designated to work with us.
Before adopting Automation Anywhere, I did not consider many other solutions because we were already working with them for years.
My experience with deployment was smooth.
My experience with the pricing, setup cost, and licensing is mixed. It is somewhere in between, not the most cost-efficient but also not the highest cost, so it can be better.
We have started exploring Document Automation. We have some in-house solutions as well for Document Automation, and we are exploring whether to use our own homegrown solutions versus Automation Anywhere for this use case.
I can give you an example where similar things like laptop procurement and provisioning come into play. Identifying where things are stuck, understanding the biggest bottlenecks in our overall business process related to IT procurement, and figuring out how to resolve tickets faster are important objectives.
The main challenge I was looking to solve with Agentic Process Automation is discoverability. We did not know where our bottlenecks were, and out of a ten-step process, which steps were taking more time was not entirely clear to us. First, it helped us see the cycle times across each of these steps and understand who all the stakeholders involved in those steps were. Eventually, this also showed us where AI can help. For example, if you need to procure a laptop, then what are the steps involved and how does AI help in connecting those stakeholders to procure the laptop faster. Those are the areas where Agentic AI helped us connect various systems and people, see where our bottlenecks are, and also prepare ourselves ahead of time so that employees do not have to wait long.
In the age of Agentic AI, I think the biggest challenges my organization faces include auditability, explainability, and observability. These are important because sometimes these agents become a black box to us. If something is a black box, we may not know why it did something or what it did. So it is important for us to know what is happening, why it is happening the way it is happening, and how to fix it. Another challenge we often face is what will happen in situations where the underlying model changes. For example, if the underlying model changes from GPT-5.5 to GPT-5.6, then the output and functioning of these models and these Agentic AI capability systems changes. We need to ensure that when the underlying large language model changes, the system's outputs and outcomes are not degraded.
AI governance is very important in my organization, especially in healthcare, where governance and compliance are the most critical aspects. Anything and everything we do has to be very secure, and we must be HIPAA and HITRUST compliant all the time. Any solution we use also has to be very secure, and that is where all these non-functional requirements like observability, explainability, auditability, and repeatability come into play. These aspects are very important.
The AI governance feature in maintaining compliance and data integrity within the organization is very helpful and plays a big role.
I do not use the Autopilot capabilities of Automation Anywhere.
I stay with them because of trust and a long-term partnership. As somebody said, it is like salsa. They know how we work, we know how they work, and we have already built mutual trust and relationships. This helps in achieving faster outcomes and quicker time to market.
My overall review rating for Automation Anywhere is nine out of ten.