What is our primary use case?
My main use cases for Automation Anywhere include both chat services within finance, invoice processing, and IT Service Management, as well as industry-related use cases, specifically focusing on healthcare. Revenue cycle management is one of the key areas where we implement solutions, and in some of my previous experience, I worked extensively with Telco organizations, improving order to activation lifecycle. Anything related to telecom processes has been a core use case as well.
A good example is when I was working on invoice processing, many tasks had to be manual. RPA used to solve some aspects, such as posting invoices into the ERP, but when it came to dispute resolution management, substantial context and reasoning were needed. Agentic Process Automation is now able to examine multiple documents at once, whether it be a purchase order or goods receipt, and it can analyze contract documentation to understand contract language, sending dispute-related emails to vendors and suppliers directly. This was one of the very first use cases where I started seeing success, and since then it has been constantly growing into other areas.
I utilize Document Automation in my current process, depending on the use case. Invoice processing is one example that heavily relies on Document Automation.
What is most valuable?
Automation Anywhere has helped me achieve my automation goals.
The cloud-based nature of Automation Anywhere CoE Manager has been beneficial. We are able to put it in the hands of business users for them to input ideas, and then for the CoE team to review the pipeline of automations and make decisions on approving the entire governance lifecycle, which has been helpful.
I use Automation Anywhere's AI Agent Studio for my automation processes.
My experience with the integration of AI Agent Studio has been very positive, specifically because I can connect with multiple systems via APIs, and the main advantage with AI Agent Studio is that I can also tap into traditional RPA workflows if I want information from a legacy system. That integration is also possible compared to other traditional AI agents.
AI governance is very important in my organization, and AI Agent Studio is improving its compliance with it. Initially, I struggled with metrics, starting from token consumption to see how exactly the agent took a particular decision. It is improving and is not fully mature yet, but all the new features regarding governance give me hope that I will soon get a complete picture of how we are leveraging the AI agents.
Compliance and security are embedded within Automation Anywhere platform, and we also have the controls to implement guardrails. This allows us to restrict PII information from being passed to the system and to the cloud, giving us good control in managing compliance and data integrity for the organization.
Document Automation impacted efficiency and productivity, as it immediately started helping me process unstructured data.
What needs improvement?
The main challenges I faced when I started using Agentic Process Automation was that earlier RPA used to solve anything rule-based, but I always hit a roadblock whenever reasoning and decision-making had to be done. The entire growth of Generative AI helped, and at the same time, Agentic Process Automation allows us to provide access to a lot of context to the agents, processing unstructured information and reasoning, so that we are able to solve many more use cases compared to traditional RPA.
The biggest challenge for the organization with Agentic AI mainly revolves around how efficiently we conduct evaluations. As we all know, whenever it comes to generative AI, there tends to be hallucinations, and we cannot be 100% sure the agent will always be right. Therefore, how efficiently we are benchmarking the performance of the agents and the overall governance is crucial, ensuring that none of the mission-critical processes go wrong due to an incorrect decision by an agent.
Automation Anywhere influenced the automation programs and the tracking of ROI.
Automation Anywhere can be improved by making it much more unified. Technology is changing rapidly, and as tech leaders, we often get lost on where the real value lies. Therefore, I want Automation Anywhere to stitch together multiple features and product lines into one single platform where I can manage the entire AI stack, rather than having multiple tools, including other AI productivity tools.
My experience with pricing, setup cost, and licensing has been mixed. Sometimes customer adoption of the Agentic features has delayed projects, specifically due to pricing-related concerns as the market evolves. Selling bots and RPA-related licensing has generally been straightforward, but Agentic-related licenses require much more due diligence, making it challenging to navigate the buying cycle due to pricing factors.
Before I used Automation Anywhere, I considered and used programmatic related automation, such as Python-related and heavy code solutions. Now, with Automation Anywhere, we sometimes lean into other solutions depending on the customer's preferences, especially if they gravitate towards specific SaaS products.
In terms of approximate time savings, traditionally processing an invoice took twenty minutes, but with all of this—not just Document Automation, but with all the other features—we are able to cut it down to less than five minutes, achieving almost a seventy-five to eighty percent efficiency gain.
For how long have I used the solution?
I started using Automation Anywhere in two thousand eighteen, so it has been almost eight to nine years now.
What do I think about the stability of the solution?
My experience with the deployment has been good, with no issues there. There has been no downtime with the platform—though every deployment has some challenges depending on how effectively it is handled in the production environment—the availability rate from a platform perspective has been very good.
Which solution did I use previously and why did I switch?
The decision to change to Automation Anywhere was initially based on ease of use and the quicker realization of value compared to other tech stacks, and this trend continues with many more AI features.
How was the initial setup?
I use Automation Anywhere's Autopilot capability, but not extensively, because Autopilot sometimes gives a skeleton, and beyond that, substantial engineering goes into building the solution. Many times it has always involved building from scratch, but that is changing. The feature around using a cloud-like interface to start building code is going to definitely start saving some time.
What other advice do I have?
Automation Anywhere's CoE Manager serves as a centralized area to track the entire initiative, including how many use cases we have and whether we are trending in the right direction or not, which has been very useful for visibility and automation lifecycle impact.
Automation Anywhere can be improved by making it much more unified. Technology is changing rapidly, and as tech leaders, we often get lost on where the real value lies. Therefore, I want Automation Anywhere to stitch together multiple features and product lines into one single platform where I can manage the entire AI stack, rather than having multiple tools, including other AI productivity tools.
In terms of approximate time savings, traditionally processing an invoice took twenty minutes, but with all of this—not just Document Automation, but with all the other features—we are able to cut it down to less than five minutes, achieving almost a seventy-five to eighty percent efficiency gain.
I would rate this review a nine out of ten.
Disclosure: My company has a business relationship with this vendor other than being a customer. Partner