Automation Anywhere Reviews

We are in the retirement industry. We use it for application-based automation. We are trying to mimic our operations to redundant business users' tasks, like clearing claims and loans.

We are using version 10.7. We are currently trying to upgrade to version 11.3.

We use Bot Insights. Business users utilize Bot Insight dashboards to get analysis.

In the 10.7 version, the back-end port for the MetaBot password has looping. This has been a type of limitation. Now, we are upgrading and integrating the Control Room, which is a validating feature.

I would like them to add OCR features. This would help us deal with unstructured data and we can get more data out of images.

I would like to see more features related to artificial intelligence. 

The latest version is more stable. We have mostly legacy applications. We do interact with Automation Anywhere support for patches and DLLs which are compatible with our legacy applications. Maybe after our upgrade, we can expect more stability.

In couple of months, we are looking to upgrade. We have set up our environment for integrating in AWS into our architecture. We are setting this up now and planning to install Automation Anywhere 11.3 to see how it is.

It is scalable. Since we have only had it one year, we are still looking into the scalabilty factor.

The technical support is good. They are knowledgeable and don't waste much of our time. They are quick in resolving things by identifying the exact root cause. 

We try to resolve the issue from our side first because we are trained in the product. It is only when we can't solve the issue that we contact support.

The initial setup is straightforward, except when setting up with cloud-based services, like Amazon. The cloud-based setup process is complex.

Automation Anywhere came in and helped set up the tool for us since it was initially new when we onboarded it.

It has saved us a lot of hours, manual effort, and we can focus on more important tasks, which helps us with our ROI.

We purchase on a bot basis. Our costs are approximately $5,000.

There are a lot of vendors out there, like Blue Prism and UiPath. However, A2019 offers cross platform and single point, which is why I would recommend Automation Anywhere.

I have been very impressed with the A2019 version. This version is phenomenal. 

Currently, we are focusing more on finance, shared services, and interpretive related ideas. For accounting and finance statistic, we have reconciliation specific processes, such as rates uploads into systems. The majority of our development is on SAP and Excel. We are focusing on the financial services area, so our use cases are more related to finance and accounting. These are the use cases developed by the development team.

We are using an on-premise deployment model.

The main agenda for using the RPA is to see how we can get the benefit from reducing human capacity as well as quality output. These two are the major focus for us and what we are achieving.

Timeline-wise, it normally takes a long time to develop. With the AA and RPA concepts and their ready-to-use components, we can deliver a bot within couple of days or within a week time frame based on the use case.

It is easy to use and developer-friendly. The development time is very short. If any system needs to be changed or any functionality needs to be incorporated to any of the technology, it usually will be a long process, like months. With this innovative tool's automation, it is very dynamic, easy to develop, and deliver results quickly. With some of our use cases, we were able to deliver within the time frame of two to three weeks, which is the biggest plus with this tool.

There are good MetaBots available in the Bot Store. We were able to leverage them, as well as established our own reusable components. Considering that, it is easy to add on any software or applications that sits in the system.

We would like to have the Excel plug-in. We have a challenge with unattended bots in development which will be deployed on virtual machines. End users ask, "Why can't we run the bots whenever we need?" The new concept for getting attended bots as well as the Excel plug-in will solve this problem, but we'll wait for 2019 version, which may be a better solution for the business.

More structured and unstructured data collection will be a challenge. While we have the IQ  Bot tool, the success rate at the beginning will be lower. It will be around 20 to 30 percent at the start because you need to train the bot at regular intervals. Of course, this depends on the data. This area could use be improvement.

The marketing strategy is directed toward end users. These leads to confusion in the organization regarding, "Who needs to develop?" Is it IT or the business? This causes friction within the business. Whatever the approach, it is important to show how to get the most benefit out of the tool. There should be a clear roles for IT and the business when using the tool. This way teams and structure can be better established in a business.

For attended automation, there are still questions to be answered: How it is going to maintain the queue? E.g., there are 20 users who want to process their own processes, and when they trigger things, how will the Control Room react? How will bot work distribution happen? These things still need to be looked into, but conceptual-wise, attended automation is really good.

It is confusing: Who needs to develop the bot? The business or IT. At the basic level, the bots can be created. However, once they are moved into production, you have to make sure that the bot is stable and running 24/7 without any issues. Therefore, a lot of care is required due to the amount of controlling mechanisms required.

There are a few problems with deployment and maintenance. When it moves to other systems, the consistency will be somewhat lost.

RPA is decently established and stable as the market is coming up with new features. However, the unstable area is more the IT part of it where a lot of research still needs to be done, especially in the case of artificial intelligence. 

When we run bots, sometimes the systems will go down. Maybe it is because the same functionality worked once or even regularly, but then all of a sudden there will be an issue. It might be a network, latency, or some other issues. Still bot stability needs improvement. These are very rare situations. It works most the time, but one time it will fail and we don't know why.

When I talk about RPAs, it is to say, "Can this problem be solved with RPA?" Scalability-wise, we are in good shape. 

In terms of AI incorporation, this type of scalable incorporation would be benefit everyone. The trend looks like we are going to reach to this point in a good manner.

We were able to put more than 40 bots into production. There are around 20 more in progress. On average, we were able to save around 25,000 hours of manual efforts.

The technical support's response is very good. They gave back solutions quickly. We haven't face many challenges when using the tool. So, we haven't interacted much with the technical team, maybe two or three instances. 

We looked into moving into RPA because it was the trend in the industry.

We do regular acquisitions which leads to a lot of duplicate roles and resources along with similar activities. With this approach, we can streamline our processes and bring uniform processes across the organization. We have been able to establish this. If we use the bots on a lot of manual efforts, it will reduces effort. Going forward, we can use the bots in a better manner by using them for the strategy of delivery and appointments.

We started this journey with a PoC. For three months, we tried understanding the tool and its capabilities. 

The initial setup is not complex for the tool setup. The struggle points are establishing the right mechanisms to identify the candidates and prioritizing things. 

In April, we started the pilot. By July, we had the software as well as the use cases that we needed. We got the licenses in July or August. It took us at least a quarter to establish systems and resources. Then, we started delivering the bots from December 2018. It took six to eight months' time frame from PoC to first bot deployment.

For our bot creation process:

A business user, whoever is doing the manual process currently, will send their request with the details of the use case. Those details will be assessed by IT and the business functional groups to measure what will be the man-hour savings. Also, is it a possible candidate for RPA?Based on those measurements, we identify candidates for RPA, then we determine if they meet the current threshold of 500 hours. If it is more than 500 hours, we consider that a candidate for RPA. Based on that criteria, we identify and prioritize it. Afterward, it comes to the development team for development.

If we had been alone during the initial setup, might not have been that successful. However, when we partnered with the other groups, like consulting firms, then we were able to get the right mechanisms in place within a short time.

We did the deployment in-house.

We have a good dashboard to measure ROI. Whatever investment, licensing, and resource costs together are put in for development and delivery, we are still at an ROI of 250 percent.

To measure ROI, whenever we get a use case from the business when we do an assessment, one of the factor that we capture is the time saving. We try to measure and apply the country rates for the locations where they are doing manual efforts. We measure the time savings by applying the country rates and deriving the hours and value.

The time that it takes to develop and deliver is within two to three weeks.

It has good licensing costs which are average for the market.

We did try out UiPath, which is a good solution. 

We also look at Blue Prism.

We like Automation Anywhere's end user experience. We thought Automation Anywhere was the better solution for developing bots in the long run. It is a centralized tool because of its Control Room. It makes sense to have a solution where everything is in a centralized repository. AA has a more future-looking perspective which will help them in the long run.

We currently are not evaluating any other vendors.

When we saw the tool capabilities, we were so excited. We tried to start using them but we needed to have the right structure and mechanisms in place from the beginning to identify the use case for prioritizing. This plays a major role. 

If you are a big organization with a center of excellence, you need to bring all the people together. The establishment will play a bigger role than just developing and delivering bots. Developing and delivering bots is a very small portion, which is doable by any individual who has a basic technical background. To be successful in your journey, having the right structure upfront will help.

We experimented a little bit with IQ Bots, but we didn't see much use cases in this line currently.

A lot of people interact with RPA and the industry is really excited about it. However, you need to pick the right candidate to be successful in your journey, along with the right framework for the development. This will give you a good output. This is what the business needs to test. Pick a solution based on the organization's needs as well as the right approach. Have an assessment with an approach framework will help.

Automation Anywhere is removing the boring, repetitive tasks from the workforce.

I've worked on different models over different frameworks. 

I do the coding as well as the deployment side. I prepare documents and the user ID. Sometimes, if the user ID is not prepared, then I jump into the process to get it done. With the technical feasibility of the document, I take the technical feasibility and do an estimation to code the bot. I configure the bot, then code it according to the entity and get it reviewed from the client. Then, I run the bot through the Hapi port, as well as different scenarios which might come up.

Once the client is happy, we have a couple of rounds of testing. We have a "You Ready" phase where we code for a few days and the client provides data. We run through the data and this give them the technical results. When they're happy, then we finally move the code over.

One of the clients that I worked with to automate their process received invoices from 12 different companies and different regions. These are the process steps that we automated:

1. We could get the data from PDFs (the invoices were in a PDF format).
2. We accepted the data, uploaded it, and filtered out some of the data. 
3. Once we extracted the data, we use it to search data on particular websites for a particular invoice number to be processed. 
   
4. We used the invoice number to process the invoice.
5. We got approval for the invoice. 

I worked on another process where it generated offer letters for different people with templates. We would get the data in a dump. We would take that data put it into a template, and then into a Word file.

With another client, we used to receive around a million files. Per day, we would have to process around 70,000 to 80,000 records, uploading data where the volume was huge. We deployed the process in multiple bots, using about 20 bots to get the process done. For this client, every second used mattered. We had to code the bot in such a way that we could save every second for them. We had to figure out how much time the bot took and the cost savings for any particular data. We used the workload feature of Automation Anywhere get it done. The manual process took around 30 to 40 people around five to six hours a day, where the bot finishes between one to one a half hours. This was a huge time savings for the client.

For the client with 70,000 to 80,000, we created an SQL database to maintain all their records using bots. E.g., if a bot failed, then another bot would retry using the same information. In the "You Ready" phase, this process was 95 to 98 percent successful. When we finally moved it to production, we had between 99 to 100 percent accuracy. This took us around one and a half months to set up. Two development levels were involved along with an architect who was guiding us.

The most valuable feature is the object cloning. If you compare it to the other RPA tools, object cloning in Automation Anywhere is the most valuable. Mainly for web application, object cloning comes in very handy. Most processes are built as either a website or as a desktop application. Out of 800 commands, around 150 to 200 will be cloned through object cloning. In Automation Anywhere object cloning, there is a special feature called DOMXPATH, where we can customize it to any level. If IDs getting changed, which happens in some cases, then we find out the pattern. Based on the pattern, we come to the particular text box and button. Then, we can use bots along with parent-sibling relationships to get things done. 

I like that it has PGP Command and it is able to connect to a Citrix environment. 

We have the MetaBot feature, which is valuable because we can read reusable components in the MetaBot, then use them as a framework.

The training and resources for this tool are very good. They are so simple that anyone can easily learn it. One of the tool values is its ease of use.

The Excel part needs improvement because we use it as a database. Right now, we are using UiPath for this feature, as that RPA tool allows us to sort, search, and filter in Excel databases. 

The IQ Bot is pretty immature, in regards to AI and machine learning. I would like them to add additional logic.

I have been using it for almost two years now.

Automation Anywhere is quite reliable and stable, especially compared to the other RPA tools. E.g., with UiPath, code that was working yesterday might not work today.

We have experience minor issues with Automation Anywhere, such as with the object cloning, where it required us to restart the machine. Once the machine was rebooted, the solution worked fine.

It takes one or two people to handle the daily maintenance of this solution.

We code our bots with tasks and subtasks in such a way that we can easily combine them when new features come up.

The initial setup is pretty simple and fast. It can be done from the web Control Room, which can be done easily.

For coding, we are keeping the information in XLS or XML files. We use MetaBot to retrieve the data. We keep our URLs in an Excel file, so when we go to our production environment that it will automatically take the URLs and implement them.

One or two people can deploy the solution easily if they understand the requirements.

In most of the cases, the product value is very good. If the infrastructure, implementation, and framework are good, then generally, the client can get a good return on investment. However, getting good resources is tough since most people have limited experience in RPA products who struggle a lot to use these tools.

The coding for Automation Anywhere is a lot easier than for UiPath and Blue Prism. I have about a year and half of experience with UiPath. Compared to UiPath, Automation Anywhere is easy to use and bot development is faster. If you try to develop the same bots in UiPath, development of the same bots in Automation Anywhere will be faster. If the client wants the bots to be developed very quickly, we will do it Automation Anywhere. However, if they give us time, then we will us UiPath. If the process is huge and complex, then we go with Blue Prism.

Another key advantage of Automation Anywhere is the object drawing feature. This comes in very handy. Whereas, if we use UiPath, we have to use the .NET code and commands versus Automation Anywhere where can use scripts or create MetaBots. 

Companies need to do feasibility testing to avoid trying to automate processes which are not right for automation. Then, they should examine which vendor is the right one for the automation process, e.g., Automation Anywhere or UiPath. They need to establish the proper infrastructure for the RPA product, like licenses and an RPA team. The RPA can be a mixture of trained professionals and people who are learning as they go from the training provided by the tool.

Version 11.3 updated a lot of features that were previously not there, like workload automation and analytics dashboards.

I am always learning new things with this tool.

Our company is an implementation partner for Automation Anywhere, and XLNC Technologies is a consulting organization for whom I work. I am the head of delivery operations.

We are working with around 14 clients who are on Automation Anywhere. We have been able to recommend and further resell starter and enterprise packs. Implementations performed by us scale up from one bot to 10 bot runners

My primary use case for Automation Anywhere would be the one which yields higher ROI , has repetitive tasks to be performed and doesnt involve fuzzy logic for sure . Cases that we have achieved success are extraction of data from PDF's , SAP automation, AD Integration, Integration with CRM and Payment links, KYC documents and web scrapping. Listed are low hanging fruits and will boost confidence of the stakeholders in the RPA deployment , recommend to look at such use cases when looking at Automation Anywhere and its implementation.

I have worked in Dubai, and we implemented RPA for one of the major retail giants. We started out with one process (bank reconciliation) and scaled up to almost 200 bots in six months. Sales was improved by at least by 80 percent. 

During the Dubai shopping festival, there were big launches that happened on the sales/retail domains. Bots compared products with other competitors, looking at the pricing ranges then locating the price for your own product, then bringing them to the home page for easy access. All these sort of changes were managed by bots. It had a direct impact on the customer lifecycle. 

In my current implementation, which is happening in India, with the poker companies, if you played poker, then you would have observed when you are trying to make out a cash-out if you don't get your cash-out in four to eight hours that you tend to get worried about the poker game in which you were playing. One of the use case that we received was to go ahead and make cash-outs almost on a real-time basis along with the tax deduction. Because in India, there are tax deductions which happen. Before, it used to take almost a day to go ahead and get the tax deducted. Then, the cash-outs were being made in eight hours. Now, it is happening in one to two hours. This boosts the customer confidence and experience. When the customer experience increases, obviously sales will increase. 

Automation Anywhere has multiple features in it. If you could look at the agnostic tools that are available on the market, almost everybody has a similar features, but Automation Anywhere has quite a secure platform. It is highly encrypted.

It is highly scalable on cloud and virtual machines. 

Automation Anywhere is a tool. It is not able to go ahead and extract the hand-written documents. Most of the automation tools that we have in the market are not able to get an accuracy of close to 80 to 90 percent on hand-written documents. So, I'm eagerly looking at a release from Automation Anywhere, wherein they could plug in an LP or look at some sort of a platform which could convert hand-written documents into a readable format. If they could bring in they own platform, this would make a big in-roads for the entire automation industry.

If it is able to integrate with any plug-ins available on the market which could read hand-written documents, that would be good. IQ BOT is one cognitive bot that they have launched, and if they could extend this particular cognitive bot to hand-written documents, then they would have an edge over other automation giants already in existence.

Recently, they launched the business analytical dashboard, which is cutting-edge. We can see in real-time the business dynamics and the way the metrics are behaving when the bots are being run. If we could have a ready-made application allowing for input processes and metrics into it. Then, it provides the approach that you should take to start the automation and when the process should expect to be completed. 

Automation Anywhere already has bots stores and digital workers being enabled on the website. Instead of putting it on the website, if a couple of digital workers could be made inherent in the tool itself, then the customer who is buying the package gets an edge over other partners or other vendors. E.g., if Automation Anywhere could package a digital worker along with my existing tool, as sort of an add-on, then that would add value to my package. Similarly, if I am from SCN, maybe an SAT module digital worker could add value.

There are hundreds of codes or utilities available. 60 percent just go down the drain and are never used, because there is no utility towards them. If they did something like Amazon, a pay-as-you-go, to get a digital worker, it would be cutting-edge technology. This would save costs for me and would optimize my implementations. The same benefits could then be shared with my customers. 

Stability is an important part of the product. It has accolades across the globe. If it did not, stable companies in the tech and finance industries wouldn't have gone for it. It's stable because it's highly scalable on cloud and virtual machines. The codes developed in Automation Anywhere don't tend to go back, skip, or miss out. 

The current version launched is 11.3. We are more comfortable with 11.3 now. However, the earlier version, 10.7 was good to work with, but since there even newer version coming, we work with 11.3 for now. Version 11.3 does have some bugs, but they are manageable because there is a good technical support team provided by the Automation Anywhere. We haven't had any major hiccups.

The moment that you start having access levels missing or security is being quite stringent about not opening a couple of bots, they are not providing admin access rights nor providing accesses to VPNs, that's when the stability issues come in. Then, we start blaming that the application stability, which in fact is wrong, because you are not meeting the prerequisites.

Other tools tend to give me exceptions when I am scaling them on the cloud. While the infrastructure might be available, systems won't be able to talk to each other. The codes tend to miss paths, and sometimes codes are not getting applied in the application. However, with Automation Anywhere, we haven't seen any such issues until now, wherein large scale developments are being found to be unstable. 

The good thing about Automation Anywhere is the website has a detailed architecture which talks about having a load balancer and how you can scale up. 

Currently, I have been working with clients in the US and Dubai through remote support. I have been developing bots remotely where I have been able to access their virtual machines. The virtual machines are separate for creators and runners. I have around six creators on virtual machines, which is good enough to start off. There are around 25 runners which are deployed on 25 VMs. This is in US. That's how we are operating them, and it's the largest setup that we work on.

In India, clients start small. They start with a starter pack, which is only allowed to be issued by an implementation partner. If you go onto the Automation Anywhere website, there is nothing called a starter pack. So, organizations will tend to start small with a starter pack, which may come in a bundle of two or three creators, one runner, and one controller. They want to test how it is functions in their organization. If they can pick up low hanging fruit, which means if they're picking processes which are smaller but high in volume, then they will have higher ROI. Then, these organizations tend to jump onto an enterprise level deployment in a maximum of six months. That is the trend that we have seen. 

However, the moment the client starts going big and is not able to get the right ROI in place, that's where they tend to start dropping the ball. Then, they start looking for another RPA software or decide RPA is not for them.

I have been receiving good technical support them, and hence I'm sitting on the sidelines. This is because if I didn't get the right support along with the right documentation and KMS, then I would not be able to fix issues first on my own. If I'm not able to do this, then relying on the technical support would be the last thing that I would do. The technical support that Automation Anywhere provides is fantastic, especially their acknowledgment and response times. Their knowledge base articles are quite descriptive. You will be able to fix issues on your own if you are quite tech savvy. Even if you are not, then they're quite detailed enough to follow.

If you don't get the right support, then there is an escalation channel available which you can reach out to. There is a community, where get solutions. There doesn't seem to be any issues with the support currently being provided by Automation Anywhere. I find them quite responsible and available whenever I have issues.

It is quite straightforward doing an initial setup of Automation Anywhere. This is because it has a well-defined installation guide which talks about how to install Automation Anywhere software. However, automation only works when IT support is quite adequate. So, if the IT and infrastructure teams at the client site are not supportive, for example, then I'm not able to provide admin rights or open security ports. That is when the entire installation moves into a messy state. There is a way to cater to such issues. Because this is a new technology coming, if you meet with the IT support team along with the CISO of that particular organization and set the right context and objective that company wants to achieve with automation, this roadblock can be fixed.

Otherwise, I haven't faced any issues with the setup.

We start off with the planning phase. Then, we go into the asset process gathering phase. We map it along with data gathering, wherein we gather data about the collaterals, the documents which are required in the process, standard operating procedures (SOPs), user login credentials, access rights, and business logics. Once that is done, we recommend a target operating model. Once approved by the business, that's when we start development. We start doing regression testing when post development is over. Finally, we do the handover the new app to the customer operations and can deploy the bot.

If I have to do a process discovery, which includes understanding the data gathering and creating an address process map, it takes a maximum five to ten business days (simple process). It takes this amount of time because the SME has to give me their time, and I never get all the information that I need in one shot since there are a lot of business and technical exceptions that I have to cater to. I have to be quite inquisitive in asking in the process, then the process gaps and the exceptions. I have to make a map, follow up with a sign off, then get it close.

A medium level complexity project can take 25 man days and a complex process could take a least a month and a half.

If I am deploying eight to ten processes in a span of six months, I need maximum three developers. If I'm developing out 50 bots in six months, then I need 20 developers.

Once the bots are deployed, I need a support team which works in three shifts. My support team would consist of a solution architect and three developers spread across three shifts. The monitoring team would consist of a solution architect and one or two IT support people. This is for a large scale bot development. If it is a smaller scale deployment, a smaller support team is needed for monitoring.

We have saved customers 400% in terms of time saved by implementing this product.

We implemented automation in an HR domain. We had 60 percent reduced error rates, 25 percent freed working hours, and the response time from the HR department was faster than before. We automated candidate screening, payroll automation and expense management. In finance and accounting, they had a reduction in 70 percent error rates, 35 percent working hours were freed, and there was improved satisfaction for the internal customers after the automation implementation. Finally, in IT, there was 90 percent is the reduced error rate, because in IT everything is binary.

The pricing and licensing of Automation Anywhere plays an important rule in the Indian market because in the Indian market $10,000 USD is too much. Hence, the pricing tends to go down depending on the customer relationship with the partner: A starter pack is $10,000 and an enterprise pack is $100,000. If you go through an implementation partner, you can get good deals. They can save some money. 

An annual contract is really beneficial for support.

I have been working in RPA for almost two to three years now. We started off with Softomotive. Then, we began working with UiPath and Automation Anywhere.

We have (or are currently working on) PoCs with WinAutomation, WorkFusion and Blue Prism.

The biggest difference between Automation Anywhere and their competitors are the spread that they have. You don't need to look out for testimonials on why to use Automation Anywhere. There are big giants who are already using it. The feature list that they carry with them makes them stand out. Their industry first bot, IQ Bot (a cognitive bot), has already won an award for the best technology deployed in 2018. It goes in and extracts data from invoices and applies machine learning if the data is not set to the right columns, putting it back to the right columns, then goes ahead and extracts data from purchase invoices. 

It's the best product that we have at this moment.

Start small. Don't take big steps if you are automating.

The primary use case is hard to say. We have used it in a number of different areas in our company. We've used it for HR, IT, and our business partners on the front lines using the application. Therefore, our primary use case is hard to say, but basically what we try to do is automate the drudgery out of our business partners' lives.

We have automated some of our HR applications, where there are certain courses. Because we're in a financial industry, we have to maintain certain certifications, etc. Some of that is automated where we help generate reports back to HR. In some cases, there are some very front-end, which are CSRs. We have processes automated for them so they have a less manual work effort. 

We have worked with our compliance area. Some of the things that we have automated there, because we are a financial industry, are political contributions where we have to be very careful. There are a pay-to-play laws in the US, so we've actually automated, which were very difficult because you had to go to all 50 states to download all the information. Trying to do it manually was probably over 8000 hours a year, and now, we're doing it with automation.

What is important to us is continues improvement and learning. We want everyone to be able to look at what they do with a critical eye towards how can improve and get things better? By introducing the solution and working closely with our business partners every time, we create a bot, then their mindset shifts. 

Now, they are looking at everything else they do, and saying, "Hey, wait a minute. Maybe I can do this in a different way." Whether it's using a bot, some other solution, or sometimes even just improving the process as it is without automation, the company in different parts of organization is really starting to adopt the idea of continuous learning.

Probably ease of use has been its most valuable feature. It's been very easy for us to use. We like to operate in a sort of a federated model. So, when we originally started, it was just our vendor partners who were doing the development. Now, we have our own set of core developers at my company. Since 2017, we now have about 23 different people, most of whom are certified developers, a lot of them didn't have a developer background when we started.

With the user interface, a lot of the parts of it I really like, but there are some things that could be made a little simpler. A little less clicking around here and dragging over there to use.

My impressions of the Bot Store are I really love the idea of it. We've actually downloaded a couple of them and used them. One of the more recent ones was interesting. It required a license key, which kind of confused us. If it's freeware, why do you need a license key? But we got around it.

What has also been nice, with the most recent bot from the Bot Store that we downloaded, it was developed by a vendor partner who we are partners with as well. So, when we had some issues with it, we were just put in contact with the developer at that same firm, then we were able to work through any issues. Subsequently, they have made updates to it and uploaded it back to the Bot Store, so it's actually great.

The stability is pretty good. In all fairness, I know we went from version 10.5, where we thought we were fairly stable, and we did have some issues. Then, we migrated to version 11.2 and encountered some other stability issues around scheduling things. But, after talking with some Automation Anywhere techs who were extremely helpful, we have recently, as of last week, upgraded to version 11.3.2.1, and that seems to have solved some of our problems.

We are still testing it. I still need to do some checkups. Obviously, I'm here this week, so I'm not doing it right now. Generally speaking, it is stable. Knowing what it's trying to do, I think 100 percent stability with everybody will be different. Architecture and environment are going to be pretty difficult, but they do a good job.

The scalability is easy. Installing license keys, bringing up new Bot Runners, and all that stuff from the Automation Anywhere side of it are easy. It's very easy. I've had little-to-no difficulty doing any of that. I have to go through the paperwork of hardware, VDIs, etc, and that's all on our side. But, from an Automation Anywhere standpoint, it's been very quick and easy to scale.

The technical support, from my experience, has been very good. Granted, sometimes there has been a little slowness, but if I deemed it critical, or whatever, then I've had conversations with our customer service manager, or anyone, and they've been able to get me with somebody in architect, which has been phenomenal.

When we did the upgrade to version 11.2, it was sort of complicated, because we had to start all over. We had to bring in new hardware and a new set of new databases, but the gentleman who helped me out was great. He fully documented the process for us, which was awesome, because I didn't have that from the original installs of version 10.5. Because he documented it, I was able to go through the upgrade from versions 11.2 to 11.3 very simply and fast. I would say the technical support is very good.

Why did we want to invest in an RPA solution? This was our first solution, so we weren't with someone else first. We did do some evaluation, looking at how and what we wanted to do with bots. Did we want to just use it purely on the back-end? Did we want to use it on the front-end? 

Some of the leaders at the time in the strategy and innovation team at our company just saw the need for it. We knew there was a lot of manual redundant processes that were just right for automation, so we said, "Hey, let's bring it on." The PoC sort of proved the concept, then we just brought it in-house and kept going from there.

The initial setup was somewhere right in the middle. It wasn't super simple. We could've probably done things a little better than we did, but it also wasn't horribly complex.

We used an integrator and reseller for the deployment, originally. We worked with HCL Technologies, but we found some gaps in our knowledge because of this. We didn't fully understand the architecture and how it was built. 

During all the process of the installs, we didn't get all the documentation. This made some things a little difficult for us, especially when we ended up parting ways with HCL. 

Then, we signed with a different vendor for both development and reselling, OPTIMIX. So, going through with the Automation Anywhere architects and getting it installed was great.

One of the first steps that we do before automating anything is we determine if it is a process where there is enough ROI involved to dedicate development resources or a bot to? Sometimes, in all honesty, the use case is we want to learn something, so we do it anyway. But, generally, we do a time and motion study, so we understand how much time we are saving somebody. We measure that. Additionally, we measure cost. That varies depending on the department that we're doing the automation for. But, we can measure that, and usually that is a straight hourly rate times the time saved.

Cost avoidance is something that we do, too. In other words, there was a process that we automated where a department was considering hiring someone just to do some file transfers and copies, etc. We automated that process, and they just didn't have to hire somebody.

It definitely saves us a lot of time and money.

It looks like it will be right around $115,000, not counting IQ Bot, which we won't renew until later.

We looked at Blue Prism and UiPath, but we felt that Automation Anywhere was a good mix of both the front-end and back-end, whereas UiPath seems very front-end only and Blue Prism is very back end only. We needed a mix for what we wanted to do.

We don't use it as attended as much. We have had a few cases where we were going to use attended, then just decided that for the rarity of those types of uses that we didn't want to have a Bot Runner just sitting and waiting for someone to chime in five times a day. Instead, we've created more scheduled. Right now, all of our bots are running unattended.

It's really easy to use. Again, going back to something that I mentioned earlier, we have a number of people at our firm who are now certified RPA developers that had no development background. They did that just by the online training in some cases. In some cases, it was the online training, as well as a three day class that we brought in-house and had taught. However, the only part where you want to get experience and learning are around error handling.

I've been in software for a long time. I've never seen a perfect piece of software, yet. I've seen some that are very good. With everything this is trying to do and the complexities of the environment, I'm going to give it an eight (out of ten) because it's very good. I think me giving something an eight is pretty high in this space.

Our primary use case is anything that is structured data and highly repetitive. We also see a lot of gain right now in paperwork, so AR, AP, and record to report. The cognitive product seems to be doing a really good job, and the pace of that going fantastic.

We have done a little bit of everything: 

• Oracle
• Salesforce.com
• Using Citrix, we did an order to cash.
• In the IQ Bot world, we seem to be doing a lot of AR and invoice-to-pay (those are the two that we see the most). 

We will be moving into medical billing next. Since we have done some really good prototypes with medical records, I think we are ready to go there next. 

One of the largest use cases that we have about a $10 billion company which does something like 20 thousand AR invoices a month. They had a solution in place. We actual replaced that IQ Bot, and we are hosting it out of San Jose right now. We are getting some phenomenal straight through processing. So, we took what would have taken 30 to 35 people and are able to do that little bit more accurately with about six.

Believe it or not, the first client that we ever did was a Citrix client. We cut our teeth on Citrix around the RPA product. It has some quirks to it, but it is becoming much better. So, we have quite a few instances of that Citrix machine in production at the client. 

The most valuable feature of our solution is the cognitive computing because that is coming onto stream a lot faster than you would think. What used to be in the middle of, "Can this be done or not?", is becoming, "It can be done." We are getting some incredible straight through processing rates with AR around healthcare and financial services. 

Integrating RPA and cognitive with other solutions is relatively straightforward. Originally, it came out that Automation Anywhere did screen scraping, but the technology is way beyond that now, where you can do object cloning, even moving into API. We have even coded a couple of APIs into the applications that most people will use. We actually do that now instead of going through a screen. 

Our impressions of IQ Bot are that it goes relatively fast at the pace that it is being advanced every six months, which is good. I believe that we have a couple of the largest installations of it in the U.S. right now. So, we are processing somewhere in the neighborhood of tens of thousands of AP and AR invoices. We actually have robotic operation centers on the West Coast and the East Coast. 

The IQ Bot has room for improvement. It is not that it doesn't do a lot of things today, it is just that it can do so much more. I think that that product is going to go crazy.

The next big thing that you are going to see is the analytics engine. I think people finally realize that you can do analytics on what the bots are doing. I think that you will see that come alive in the next six to nine months.

If you write them correctly and have good architectures internally, then scalability is relatively straightforward. This also applies to reliability. However, you sort of need to write a few before you get to this point. As more of these go to production, these are going to be the two big drivers.

Probably one of the faster scales that we have done is we did a prototype with IQ Bot that had about 500 invoices and a 1000 vendors. In about two and a half to three months time, start to finish, we went from the original amount to 15,000 invoices and 2500 vendors. I would suggest in that 90 to 120 day time frame, if you have done things properly and your architecture is good, that you should be able to scale at any pace that you care to.

We have seen both straightforward and complex initial setups. If you are behind a firewall, it is relatively easy to setup. When you are hosting the application, that is when it is a little more complicated (we actually host applications for people). You need to talk to the IT department, as you are actually outside of their firewall, then you are coming back inside their cloud. 

Depending on the architecture, it can be relatively straightforward, and it can be put up in just a day or two. Whereas, if you are hosting it and coming back in, it is a little more complicated.

What we did was we took the Automation Anywhere ROI calculations and added quite a few things to it. We actually ask 15 to 18 questions around process time, number of systems used, how people, and how many hours. Ultimately, we can stack rank those ROIs on a process, then we typically try to start with the largest ROIs first. So far, with what we have done, we have seen has been relatively close (as it is almost impossible to hit the ROI exactly), but the ROI on the product is good.   

No one is balking at the cost. They are market rates and will change, but no one is saying, "That is ridiculous or impossible." They are saying, "Show me the ROI and prove that the cost is accurate."

One of the things that will be good for the community will be to get the free version to download, which came out just a month ago called Community Edition. So, you can actually download it and try it for yourself. 

Ease of use is something that you need to get trained on. Once you have gone through the proper training, it is relatively straightforward to use. There is quite a bit of online training on their website, as well as in the Community Edition. After 30 to 40 hours with it, you should be pretty good with it.

One of the promises that we made to Automation Anywhere when we started was that everyone in our company would be trained and certified. Even as a Board Member, I am certified. 

Bot creation is relatively straightforward. Probably within 20 hours, you should be pretty good at it. There are some nuances around it though that I think separate first time developers versus seasoned developers. What you will see in the marketplace is how easy it is to develop one compared to how easy it is to develop one that runs in production 99.9 percent of the time.

Across the board, I would give it a nine out of ten. There are some areas that they can move up into, but they are working on it right now. I am assuming that everybody is working on it, so we will stay in that nine range.

Go do your homework. Call some of the larger clients because they seem to face the largest problems. Automation Anywhere has five to ten gigantic accounts that you can just talk to. Ultimately, I would just be careful in the way that you implement it. It is like any tool, if you know what you are doing, "Great," and if not, then you can make a mess. Therefore, be careful and know what you are doing.

We use it for backoffice processes. We are part of a shared service, so we do billing, credit, collections, and those types of processes.

We have done close to 50 projects in the last three years. AA improves our speed as far as moving files goes. It also helps in getting people prepared to work on a daily basis. We have a lot of jobs that run at night so that when people come in in the morning, everything's ready to go for them.

For example, we have two very separate systems which need to talk to each other. It sounds very simple, but we have a job that takes a file from one system and loads it into another. But it has to do that close to 1,500 times in the middle of the night. It's a job that used to be done manually by 150 people in 150 different markets. Now, it's something that's done during the night, and when they come in it's all completed. So it was a very simple task, but there was a lot of volume. It has saved everybody a lot of time and it has saved a huge amount of manpower. It saves us thousands of hours a month.

It's also very useful as far as interacting with employees. Employees can kick off jobs on an as-needed basis rather than scheduling them. It's always very helpful to have a tool that's interactive with the employees.

One of the most valuable features is object cloning, the ability to get into websites and do things faster.

I also like its usability. It's pretty easy to learn, pretty quick to get things taken care of with it. Our average developer is up within 30 days. We have our own training program. We don't use anything from AA for training. We've been doing it long enough that we created our own.

In addition, the user interface is easy to use. It's fairly user-friendly when you don't know anything about it and open it up for the first time.

The scheduling is a little difficult at times. Rather than setting up individual instances, it would be really nice if we had the ability to set repetitive jobs easily. Right now, if you want a job to run every 15 minutes, you have to schedule it a lot of times. I'd rather have the ability to just say 'run every 15 minutes.'

There could be some improvements made in the Control Room. I really like the concept of the application that they've got, so you can now access the Control Room from a mobile device, but there's only a lot of potential, there's not a lot of functionality there yet. That would be a great place for it to be able to expand, so you could have full functionality of the Control Room through a mobile device.

And in general, I would like to the solution to get into more machine-learning/AI. I know that the IQ Bots are looking to go there, but there's definitely a lot more potential there as well.

We haven't had any problems with the actual application going down. We really haven't had any issues. We have issues with third-party products going up and down, but we've never had AA just stop.

We're a small user, so we haven't had any problem with scalability. We've talked to a lot of people who have used it on a much larger scale. For us, it's very easy to add TaskBots. We're more than satisfied with the scalability.

We're always looking for expansion of the solution within our company. We use most of the commands available. There are very few that aren't applicable to what we're doing. We're always trying to get the solution into other departments within the organization itself. We use it a lot within the shared services, the area which we own. But outside of the company, we have several projects that are not within the shared service and we're always looking to talk to the other departments and get them involved.

We have 25 bots, meaning 25 licenses. Our core team that does development has four people plus a project manager.

Tech support has greatly improved. Three years ago we didn't necessarily have the best experience, but over the years they've gotten better.

As a side note, we go through a third-party first, through ISG. That's through how our service model works. Then ISG gets AA involved if they can't answer the question. We don't necessarily directly contact AA. Most of it's through the third-party provider, and then AA eventually.

ISG is great. We really haven't had a lot of problems. When we implemented version 11.3, we went so quickly - we went with it right when it came out - that there were a few questions that ISG wasn't aware of yet because it was so new. We went to AA through ISG and they were able to answer the questions right away. But for the most part, ISG is very on top of it, and we don't need to engage AA.

This is our first go in the RPA world. We had an internal resource, somebody within the company who had seen Automation Anywhere used in other companies. That person introduced us to the concept and, from there, we did some research and saw that that's where the market was going.

The initial setup was pretty straightforward. I did it myself and I'm not an IT guy, so it wasn't super-difficult. It took me a couple of hours, including configuration. Our IT team set up the actual virtual machines themselves, but the installation was done completely by me.

Our implementation strategy was that we started super-small. At that time, we started with five bots and the Control Room. It wasn't overly difficult. I just followed the documentation provided to install the service and configure everything.

Our initial project had to have an ROI, and that's why we started small. We have never not had a yearly ROI. We've always greatly exceeded the cost of, or the investment in, the tool. In terms of how much it's saving us, I don't want to be quoted on the exact amount, but it's more than millions of dollars a year.

We purchased AA through ISG and have done both annual renewals and a multi-year renewal, the latter recently. It has been very easy to add bots on an as-needed basis.

There are no additional costs from Automation Anywhere, but there is obviously the infrastructure costs for the VMs.

We definitely evaluated lots of different options at the time. Of course, it came down to Blue Prism or to AA. We went with AA because of the feature set, the way that the development was laid out. And support was also a factor. We were much more interested in States-side support.

Any company in today's environment would be foolish not to implement RPA. There are definitely different types of use cases where you could spread it out amongst the organizations and let them do their own thing. We have chosen to keep it centralized and have been pretty successful in doing it. But everybody should be using an RPA in some capacity.

I would absolutely recommend specifically Automation Anywhere.

Our experience with the solution has been a ten out of ten. I can't speak to the other tools because we haven't used them, but we're very satisfied with what we have.

Our primary use case is banking/financial. Processes we have automated include loans, ledgers, mortgage loans, and even some of the record management systems.

In some of our use cases, people were spending more than three hours per day just generating reports. And then we created an automation for this and it reduced the time to 30 minutes. It improves employee productivity so they can use their time in other areas.

I like the way it works with structured data in the back office and the way it does repetitive work.

In terms of ease of use for developers, we're able to create reusable components. We don't want people to have to rebuild things from scratch. In this way, developers can take the reusable components and complete their development processes more quickly.

The bot creation process is pretty straightforward. Anyone can go in and learn it easily, and then they can build a bot. I like it.

When it comes to integrating the solution with other applications, there are some challenges. For some third-party solutions, there are no direct interconnections. For example, there were no direct connections with SAP systems. So, we had to create connectivity between Automation Anywhere and some third-party solutions. They have now improved that situation a lot and we can connect SAP and other systems as well.

If they want to sustain their position in the market, they have to be flexible, working on how we can integrate with third-parties, working on a machine-learning product. People are expecting that and it would be really helpful.

From the IQ Bot perspective, frankly speaking, they still have to improve a lot. I got IQ Bot training in San Jose. My expectation from a straight, technical, architectural point of view was that I would be able to create my own algorithm and integrate it. But with IQ Bot, I am not able to integrate anything. It is already well-defined, so I have to use that particular option only. I know I can not go with any other machine-learning platform. I hope they will be coming out with version 12 where we can integrate it with Python algorithms and other stuff. It might only be in the future, it might only be on the roadmap. But as of now, it is lacking a lot in that area. We are expecting, for most of the documentation, things like tags, that there would be a checkbox option. That's lacking in IQ Bot.

The stability has increased a lot. When we started with version 10.2, there was a lot of instability. There was no way we could keep the bots active, there were scenarios where it became disconnected. There is also the code deployment perspective and a lot of other angles. People are always only thinking from the business perspective, but as a technical architect, I think about operational effectiveness and how they can improve the product's maturity.

The stability has improved a lot.

However, when upgrading, they changed their internal architecture. They moved it to a JT Java platform. When moving, some of the existing features did not work in the new version. It might be that they have to improve their regression testing to improve clients' satisfaction. It can happen that what is running in production currently, if I move to a new version, suddenly is not working tomorrow. People are not happy with that and say, "I want to roll back to the older version." They are not able to use the new features.

When moving to a new version, they have to think about what features people are using and what kind of impact there will be. Small business users will be fine, those who have ten bots or 15 bots. But there are organizations like mine that have around 700 to 1,000 bots, and that makes changes very difficult to handle. It could be that 10,000 tasks are using something and if that thing is changed it will be hard to update. I might have to spend a year on that. People will never accept that.

Scalability-wise, they have increased it a lot, based on the clustering method. As a technical architect, I'm going with always-on production and data centers. That means that if any data center goes down - there is a natural disaster or something else that happens - how do you make it such that you can bring up another data center? I'm coming up with a design for that based on the latest version, version 11.

The initial setup is very simple. It's Windows-based and it's a straightforward installation. We used to say they need to come up with a silent installation option, with the previous version. But now, with version 11, they have introduced, even at the server level, a silent installation. That means we can make it automated instead of manually installing it.

We measure the ROI of automated processes by how much of a benefit we're getting from it. We look at how much time it takes and how many robots we're using and we include the licensing and operations costs. Finally, we take into account how much faster the performance of the bot is, compared to how long it took to do the process before automation.

We have saved time and money, but when people think of going with RPA they cannot expect that they will immediately see ROI. They have to sustain and increase the RPA options. They will have to spend a minimum of one or two years increasing their use cases for automation. Then they will see a good ROI. They should not expect, within three months, to say, "Hey, I have automated, where is the ROI?"

All organizations have a certain strategy or checklist. In this case, management will think first about licensing cost, about the total cost of investment. After that, they will think about the product's features and functionality. They will also look at support. They will consider the use cases, the current processes they have identified already, and based on all that they will decide whether to go with Automation Anywhere or another product.

In terms of our bot creation process, people come to me and say, "I have a process. How do we automate it?" We need to understand if it's a cognitive use case or a straightforward use case. If it's straightforward, we tell them we'll use this product and build it for them with four to six weeks of development. Then it can go to production. If it's cognitive, then we really need to understand it better. We need to use a third-party product, like Kofax or maybe an IQ Bot if it is fit for the scenario. Based on that, it takes some time and then we'll move it to production.

We have a process architecture review committee where we review all the processes. We cannot blindly go forward with all the processes that have scope for automation because it's all licensing cost. We need to think about whether we can automate a given process with any other IT automation solution, like scripting or macros. If that is not possible then we have a fit for RPA. Then we have to go through our checklist, walk through the use cases, and look at the percentage of the automation scope: Is it a 100 percent automation scope or 80 percent or 20 percent? We need to to know if there is any manual validation or manual intervention required and how that is handled.

Initially, we failed with the Citrix automation where we have a lot of use cases. We ran into a lot of limitations with Automation Anywhere in version 10.5. But with version 11, they have AI Sense which we can use for Citrix applications. We are currently exploring this option.

I have taken courses at the Automation Anywhere University and I have advanced professional certification from Automation Anywhere, which I completed for version 10.5. I'm also doing it for version 11. I also have an official certificate for IQ Bots.

At the moment, for us, everything is on-premise. We're not ready to go with cloud. So we have to build our own platform. We have to build our own bots.

I would rate this solution at seven out of ten. They have to improve on the product's maturity level. When they are introducing new versions, they have to conserve the existing commands and features, so that they work when we move to the new version. And they also have to come up with more flexibility, so their solution can integrate with our scripting and our own algorithms. That will make it easy to convince our business areas to increase the adoption of RPA.