UiPath Reviews

Right now, we have nine production animations.

Our first use case is to convert video to audio and then upload it to a legacy system. We have another use case that takes an Excel spreadsheet of 2,000 rows and searches one row at a time on the federal court website to get the results of our cases. 

We also have another use case that generates an SAP report daily, downloads a report, and feeds a Power BI dashboard. 

The last use case monitors a client's email, and as soon as there's the form request, we open the request, and we add it to their system automatically.

Thanks to COVID-19, since everything went digital, it added a lot of manual work to employees. It helps reduce manual labor. It just saved time. The robot now does the amount of work that got created due to COVID. It went back to normal in terms of workload now that we can get help from UiPath.

I love the REFramework and the Azure Queue and Orchestrator. Those are great features. They all work well.

The UiPath community is awesome. That's one of the reasons I use UiPath. There's so much stuff available everywhere. For example, if you go on Google and search for a subject, you are going to find information on either YouTube videos or community sites. There's just a lot of information available about everything. Also, a lot of super complete documentation is on the UiPath website. Any question we get, it's super easy to find technical information on that subject.

I've used the UiPath Academy courses. It has helped me start very quickly since they're all free, and there are many courses to choose from. When I started, I used my first two weeks to just do Academy training, and that's how I started. They have a complete curriculum. It's pretty nice.

The speed of upgrades is really hard to follow. It's great that there are continuous improvements just for us; however, it's just a bit hard to keep up with all of the changes.

I wish I could use the SaaS version. Our only con is the SaaS offering is not accredited in Canada, so we're working on that. That's not a UiPath problem.

I've used the solution for about two years.

The stability is awesome. It is very stable. It recovers itself. It doesn't err. It's very easy to carry on, and exception handling is very easy to manage it. I have no problem with it.

It's super easy to scale. I'm sure there are always ways to improve, and we're a small group, so I only have six bots in production right now. However, I already have some processes that could be scaled. They're using two robots now, yet I could easily put that to four or five if necessary. If you create workflows correctly, it's super easy to scale.

Technical support is pretty good. I've used it a couple of times and had no problems.

Positive

I did not previously use a different solution.

I was involved in the initial setup, and it was straightforward. 

The installation instructions were pretty easy to follow. The orchestrator was pretty easy to set up, as were the studio and the system. There's nothing complicated. I find it's just a regular install.

I was given the license information and just went along with it. We are a small department. I was in a team of one. We didn't have a real strategy. We just had to start somewhere and do a proof of concept and move on from that. That was my strategy.

We did not use an integrator, reseller, or consultant. It was implemented in-house.

I have witnessed an ROI. For example, when converting video to audio and uploading it. Each video was taking about 30 minutes to upload manually, and we produce about 280 videos per day. The robot does it in four minutes right now, and it does it across Canada. Three bots do 280 videos in 20 hours. The ROI there alone would be huge.

I don't deal with the pricing aspect. That said, I don't see it as a problem. I see that it brings a lot more value. We save a lot more than the cost of their license.

I touched Power Automate a bit. I heard about Blue Prism. However, I did not really look into it. We just started with UiPath. When I joined them, my company had already purchased UiPath. So for me, I just started with UiPath.

We do not use the AI functionality just yet, although we have plans to in the future. 

I would for sure recommend UiPath. I haven't really tried other solutions. However, I've heard that people were trying them and returning to UiPath.

I'd rate it a ten out of ten. The product is awesome. The support is great. The community is also huge. It's super easy to find information, find support, and find stuff. I find it very easy to start using the product and can show value and benefit to clients.

We use it for everything. We use it for very simple things, such as moving data around in Excel, and bigger things that include using more advanced technologies, APIs, and some of the newer stuff that UiPath has had, such as Action Center, etc.

We have not yet used its AI functionality in our automation program, but we plan to do that in the coming months.

The company is growing so fast, and we've been able to use automation to keep up with that pace. We've started rolling out to individual citizen developers. We're trying to change the whole company in terms of the way people work by using this technology.

The way they stay on top of Orchestrator is really helpful because that has been the platform that controls everything. The dedication to that is pretty helpful. Every time there's a new release, it seems there are more and more features. It's also not hard to learn.

When it comes to the UiPath Community, everybody is helpful. If you don't know how to do something or you want to learn about something, it's pretty easy to connect with other people or talk to people at UiPath to get that knowledge or learn how to do something. You can also just point somebody to UiPath Academy. They go from knowing nothing to being pretty good with things pretty quickly.

One improvement that could be made is in terms of keeping the documentation updated. Some of the online documentation is outdated. Because things are always changing, it is understandable that information becomes outdated pretty quickly, but sometimes, when you want to go use something, deploy something, or troubleshoot something, the documentation says, "Do this," but what it says to do no longer exists.

I have been using UiPath for a little over four years.

Its stability has been really good, especially lately. It has always been relatively stable, but as they've added things, it still seems to be pretty stable.

Its scalability is pretty good. Orchestrator is able to handle whatever you're throwing into it. If you need to add a bunch of processes or jobs, it's pretty easy to do. The only thing that would slow that down is if you need new servers or something from IT. So, from the UiPath side, it's pretty easy, but there are other variables.

They're pretty helpful. If you submit a ticket, somebody reaches out to you pretty fast, but usually, I'm able to just reach out to our account managers and get help within a few minutes. I'd give them high remarks. I would rate them a 10 out of 10.

Positive

UiPath was pretty involved with it. So, it was pretty straightforward. We're about to move to UiPath Cloud though, and our account managers are pretty involved in that, but it has been pretty turnkey and straightforward.

Our implementation strategy was to get a platform that works and start building things. I'm not sure of the overall strategy. Some of the decisions were made before I got here.

We worked pretty closely with UiPath. They've been great and pretty helpful. 

We have definitely seen an ROI. I don't know any of the specific metrics per se, but I know the values out there. We're just getting hammered with use cases everywhere. We did something literally last week that took us three days to deploy, and it saved a team 100 hours of work.

I came to the company as they were just starting to use UiPath. I don't think they evaluated other options.

My advice to others evaluating the solution would be to compare what UiPath is doing to all its competitors, and none of the competitors would scratch the surface of what the offerings are.

I'd give it a 10 out of 10. I liked UiPath so much that I went to get a job strictly in RPA. That wasn't directly UiPath, but before I started where I am now, I was working with UiPath just a little bit, and then I was like, "I want to pursue that for long."

I'm a web developer. This solution is used in the backend and frontend of my organization. 

The solution is deployed on cloud through AWS.

It doesn't take a lot of effort to understand UiPath's features. It saves time, effort, and increases performance.

Automation has minimized our on-premises footprint. UiPath speeds up and reduces the cost of digital transformation.

The solution has reduced human error due to the automation and scripts.

The solution is deployed across multiple locations. We currently serve between 50 to 80 clients. They are mostly small and medium enterprises.

UiPath has a very good user interface, and the automation is a great feature. New users can easily understand UiPath.

I would rate the ease of building automations and using the solution as nine out of ten. 

We use UiPath for automation and architecture design in the backend. It automatically generates data for us, so it saves resources and the work of extracting data.

UiPath enables us to implement end-to-end automation. End-to-end automation is critical for us. For the past 20 years, we have been dependent on it.

There's some latency when changing from one page to another, so that could be improved.

I have used UiPath for one year.

The solution is very stable.

It's scalable.

I would rate this solution as nine out of ten. 

UiPath's technical support is very good. When we have required information, support is available 24/7.

The setup involves load balances, and UiPath is involved in the automation and architecture of the backend. It took about one month to understand the automation feature.

We have a team of 40 people for development. Only 10 people were necessary for deployment. Some of their roles were backend, frontend, architecture, and AWS.

The solution has reduced the time our employees spend on certain tasks because of automation and scripts. We save about 60 hours per month.

We save approximately $1,000 per month with UiPath.

The solution isn't very expensive.

My team evaluated other options. They chose UiPath because of its community and positive reviews.

I would rate this solution as nine out of ten. 

I would recommend this solution because of its great technical support and its automation features. UiPath also has a great community.

The beauty of RPA is that it has many different use cases. We use the product as a standalone and as part of a bigger solution. Obviously, the tool itself is designed to automate activities that humans would do. But as the tool is getting smarter, we are able to do more types of activities.

There were two big challenges in the early days. One was the complexity of the rules that you needed to adopt for a particular task and the other was the type of data that was being used as part of the business process. As we've gone along, the product has evolved and allows us to do more of the business process.

We are able to offer more automation to a business process than we could before. It's all about the business process and how much of that can be automated and what still needs to be handled manually. UiPath gives us the ability to do more with automation and need less human involvement.

We used UiPath to automate processes for a good cause, free of charge, during the COVID pandemic. We used it to build automations that would take care of activities that nurses and doctors were having to perform, to free up their time to treat patients. Reducing the admin they normally might have had to do, rather than being front-of-house and dealing with very ill patients, speaks volumes for itself. We freed up time for our nurses and doctors to treat very sick patients.

When it comes to reducing the on-premises footprint, what we've seen over the last two or three years, with the pandemic, is the move to the cloud being accelerated. The whole digital agenda was accelerated. What we're seeing now is that most of the organizations we work with are more open to using a cloud solution. However, it depends on each organization. Some of our government organizations are not allowed to use the cloud. They have to use on-premises solutions because of strict data rules. But more companies are now open to using the cloud and that has an impact on the total cost of ownership. They're not having to invest in services and that helps reduce TCO.

AI models are now being pre-trained. There's a risk there because, obviously, you are using someone else's data and someone else's bias. But if you put those aside and use the pre-trained models, it's going to reduce the time to value because you're not having to train models. You've got something that can be run out-of-the-box. To be honest, we will always build our own models rather than exclusively use UiPath's. For instance, they have an email reader. We tend to use their model in conjunction with our own and that has helped us reduce development time on our AI models and our training.

On the subject of human error, anytime you ask a human to do an admin activity where they're typing in data, there's an opportunity for human error to creep in. And that can cause catastrophic events, downstream. Where you have an automation that is guaranteed to enter data that is 100 percent correct, you're always going to reduce human error.

Automation can also be used where a decision is made. Sometimes, human decisions can be fallible. If you have a decision being made by an AI model, it's not going to have human bias. It can have other biases built-in, but you can see an improvement in some decision-making as well. What we have to understand here is the introduction of AI into RPA is still in its infancy and there's a long way to go in this area. But there are definitely improvements being made by the introduction of artificial intelligence. It depends on how you use it and how smart your understanding of it is, as well.

Automation is all about trying to reduce the touch time of a human in a process to free up their time. If the automation is implemented smartly, it does so. If it's poorly implemented, you can end up with people being given so many exceptions that the bot actually becomes redundant. But as a rule of thumb, of course, it does free up peoples' time.

Similarly, if it's deployed smartly there are some benefits in terms of cost savings. However, we're not using it because it's saving costs. We are using it for our staff who don't want to do manual activities. It's not just about dollars and saving time. It's also about our staff and not having to do menial tasks, which can be demotivating.

It's not about the tool, it's about the business process and which component works best. The most valuable feature really depends on the use case. My favorite now is Document Understanding and the AI models that are being pre-trained. That's allowing us to do more, out-of-the-box, without having to do custom programming.

When I ask my developers which product they would like to use when we are doing RPA, they always say UiPath because of its closeness to Microsoft's .NET. Again, what UiPath are doing is enabling more out-of-the-box functionality without having to do customization and coding. The developers, and I, would say that it's getting easier to use for simple automations. It still requires planning and thinking for more complex automations, but you are able to do more with fewer skills. 

UiPath has something called Citizen Developers, which is where they encourage people to build their own robots and the functionality available is greater. Whether that's a good thing or a bad thing is a question for governance.

UiPath's user community was one of the strongest aspects from an adoption point of view in the early days. The forums, the community, were always something that they invested heavily in, and that was a smart thing to do. The community is still strong. They have a Community addition and the feedback we get from some of our customers, when they first play with this, is that it's a friendly community. People are encouraged to play with UiPath, and if they get stuck, the community is willing to provide them with advice and guidance in a friendly way. Not all communities are the same, let's put it that way.

We actually teach the UiPath Academy courses. We are a UiPath-accredited trainer, but we have, at times, with smaller organizations, suggested that they take some of the training themselves. It's another great area of UiPath, in addition to the community. Their training is available free of charge. That has always been a strong point as well.

Their licensing is poorly constructed. It's too complicated and not well thought out. They also outsourced their support model, which sadly has become less friendly and more automated.

It's very much in the early days, but another area they can continue to look at is bots building robots—the ability to take a task capture and turn that into an automation. There are always concerns over that: Who's going to police the police?

I've been using UiPath since before it was called UiPath. It was previously called DeskOver. I worked for Capgemini and I worked with the product when the company consisted of 30 people in a room in the very early days. Capgemini adopted DeskOver and then it became UiPath as a growth partner. So I have been using it for coming up on 10 years.

UiPath is stable. Like any product in the early days, they were updating all of the time. That used to not sit well with our bigger customers because of their update plans. As the product has matured, it's become more stable and the release plan has become structured. You now get four updates a year. It has become more mature.

It's scalable.

The environments that it can be used in can be very complex, such as "multi-tenant," where a tenant is the protection of data from Orchestrator, where you have segregation or air gaps. We've done some projects with a defense ministry that are very complex and we've done some very simple ones.

Most organizations start small. There will probably be a pilot with a cut-down version of what is needed. As they grow and scale, they will invest in more tenants and in more infrastructure and more components of the platform.

I would rate their support at seven out of 10, but the dial is going the wrong way. Because they've outsourced their support, you're no longer dealing with the original product team. I had a very good relationship with the product team where I could almost pick up the phone if there were some proper technical problems.

Now, that is gone and we're having to deal with a third party. As UiPath have grown, they've needed to focus on their core areas and outsource other areas. But it's getting harder to get good quality support because my customers are no longer dealing with them directly, they're having to go through a third party.

Neutral

We started off with Blue Prism, which was originally designed, and still is, for back-office processes. What we used DeskOver for was front-office automations, almost like "your little buddy who sits on your desktop." As UiPath grew, it started to erode the Blue Prism market because it offered the ability to do both front-office and back-office automations. That's why it became number one. It recognized that there were two distinct markets: front-office and back-office. Blue Prism always has done back-office, and successfully, but never offered the ability to do front-office. Customers didn't want to invest in two different technologies as that would be too expensive. That's why UiPath stole the match and has never looked back.

Back then, 10 years ago, there wasn't much else to choose from. It was really "macros on steroids." The market was very different. It was Blue Prism or some of what were very young companies. Automation Anywhere was just coming in as well. 

But at Capgemini, we recognized UiPath, and we liked Daniel Dines, who was their CEO at the time and is now Co-CEO. We liked his enthusiasm and we liked the price point as well. They were from Romania, they were cheap, and they were eager. We were able to work with a company that, at that time, was very competitively priced. And we were able to get them to start building stuff that we wanted as well, as Capgemini was one of the big six. We had a lot of control over their roadmap in the early days.

Each organization we work with will have different design principles. As a general rule of thumb, we encourage investment in the cloud license model because it reduces the time that it takes to actually get the implementation up and running, and it simplifies things when it comes to TCO. However, there are organizations that are constrained by security, and therefore they can't go down that route.

I'm an architect by trade and my job is to oversee complex delivery and deployments. It's all about the architecture. An architect needs to work with the client, in the beginning, to come up with a plan and a solution that's going to be fit for purpose. In addition to architects, you need a project manager, some engineers to actually do the implementation, and you're probably going to need a tester to test and commission the environment.

The initial deployment has gotten better. When Orchestrator, which is the main administrative console, is built-in as part of the cloud offering, rather than having it locally, it makes a big difference. The cloud has made the time to set up a pilot and deploy into production much shorter than it used to be with on-premises.

Most IT projects probably take three to six months to deliver a success. Whether or not the payback happens straight away depends on the investment costs. The benefit of UiPath and RPA is the rapidity with which you can get automation into production and beginning to pay back. RPA has always been liked by people in the business because it is a rapid deployment rather than something more strategic that can take one, two, three, and sometimes, five years if it's a very large IT program of work.

ROI isn't just about the dollar. It can be other things as well. If a program of work with RPA and UiPath is deployed smartly, you can see a decent ROI. But that ROI also depends on the declarations by the business involved. How many times do they execute the process and how long does it take to do it? How many exceptions are there? How many people are involved in that process? 

Business often thinks that volumes are higher than they actually are, and they may forget that other people may have to be drafted in if there are peak periods. Due diligence in building a business case is important. What you also need to do is revisit it six months later or a year later: What did we declare? What did we actually hit? Did the bot fall over because of X, Y, or Z? It is a continuous improvement process, as well as ROI. If you get the two working correctly, you get a stronger ROI.

What we find is that some businesses who don't use our consultancy skills will try and do it themselves, and they will come across some of the pitfalls that we are fully aware of, but they may not be aware of because they're learning, and that has a big impact on ROI. If they try to go after a business process that's very complex, at the beginning, without the right skills in hand, they can suddenly find themselves in a downward spiral from a development point of view, where costs and time are overrunning. Before they know it, the declaration they set has passed and they've not been able to hit those targets. When that happens you start to see an erosion of confidence from the business side as well. It all comes together. It's all about strategic understanding and technical know-how.

We now have the challenge of cost. It depends on what you want to do. If you want a very simple robot to execute normal front-office transactions, I would recommend Microsoft Power Automate because it's free. If you are investing in more complex things, then I would always recommend UiPath. 

UiPath is starting to price itself out of the market. It's getting expensive. That's fine if they continue to push the envelope of what they offer, but it's all about perception. And the perception over the last two years with my customers is that it's expensive. I trust it, but it's expensive. 

We're seeing year-on-year price increases. There's a price point they're getting to that is about the value and they need to be very careful about that. 

UiPath, as an organization, has changed massively. When it IPO'ed it was a Romanian company and very much a European company. Now, it's an American company with American values, and I think there is a misunderstanding of the European market compared to the American market.

In terms of implementing end-to-end automation, it depends on the process. We can do more with the advancements that I've mentioned, but every business process is different. I always say to my team, "This isn't about the tail wagging the dog. It's not the technology, it's the business process, and whatever tool is right, whatever fits." There are still exceptions that need to be handled by people so we're probably not there yet with end-to-end automation. 

Most good-sized business processes have business exceptions that require involvement from people. It's what we call "human in the loop." Two or three years ago it was 80/20 between automation and human involvement. We're now probably at 85/15 or 90/10, with more of the process being automated without human involvement. That's because there are smarter bots using AI—the brain—to execute automation tasks that previously had to be handed back to a human for decision-making or some other activity.

On its own, UiPath doesn't speed up digital transformation. It's a tool. It helps, it's a contributor, but as a standalone, it doesn't. It needs other things.

I had a quick look back at what their offering was in 2019 and what it is today. That's a good way of looking at how well they've listened to their customers. They've been smart in not just providing RPA. All their components now are far more than just RPA.

They've recognized one of the biggest areas is process identification, that whole journey of identifying an opportunity and taking it through the life cycle, with things like Automation Hub.

Generally, the response I get from my customers is they're impressed with the number of solutions that are available under UiPath's Enterprise platform.

We implement automation for clients to create savings by cutting the number of FTEs. We've used UiPath for various kinds of automation, including mainframes, browsers, Excel, account payables and receivables, fixed assets, healthcare projects, HR projects, reporting robots, and IT services projects. 

Recently, we did a massive US taxation project that spanned eleven months and covered enterprise and individual taxation extensions. It was a huge project that yielded a lot of savings. 

If I want to leverage a specific UiPath use case, I build small use cases around that particular feature and try to envision a product out of it. I've had several hackathons and general discussion calls because I'm a solution architect. Everybody wants to work on apps, and UiPath is comparable to the blank canvas apps that Microsoft PowerApps provides. 

When we had an automation program that involved 200-plus automations, we created around 100-plus libraries, saving us thousands of hours of development time. UiPath is designed to save time. The object repository was liberating because it enabled us to move from simple to extendable libraries. UiPath's apps increase our business by helping us leverage the UI layer in a way we couldn't in the past.

It gives us the ability to share data between systems in healthcare applications.. However, it's still tricky because so many system controls are in place. That's not a limitation of UiPath per se, but every department has restrictions on passing data to other departments. They have their own due diligence in place, limiting data flow from one system to another. UiPath gives us the fluidity and freedom to do it, but the limitations within each domain often get in the way.

Let's use claims data, for example. The data regulation team won't be too keen on allowing the marketing department to use data from the claims division to generate new business. The data flow from one department to another isn't that fluid. Organizational controls rather than system controls bind it.

We should look at each separately in terms of AI and machine learning. If we want to do data analysis, we have to call an inverse Python script, which is a little difficult. However, we can host our own model, and that's good. The ability to use that opened some doors. 

At the same time, it's helpful to have out-of-the-box features like Document Understanding and an ML passer there. The integration is quite fluid. We can directly call a Document Understanding model and then give it to ML passer and then get the results out. It's smoother for integration. The client has to focus on one particular software or multi-stack that they're comfortable with. UiPath has opened some opportunities in that sense. It made life easier because the capability is sitting inside the platform itself. 

UiPath is a separate solution, but it can talk to other services and doesn't restrict you to the passer, but that's how the ML features within Document Understanding help us. Custom model hosting and the AI center also help. We don't have to host the custom model somewhere else and call that service then pass it and do the post-processing within the system. It isn't a third-party service, so we know it's sitting within the system. If any issues are also there, we know where to diagnose and deposit them.

UiPath Orchestrator is a treasure, and UiPath Studio includes various packages to integrate with solutions like Salesforce, ServiceNow, and Excel. They also have mainframes, web automation, and the API package. 

Orchestrator contains a lot of useful apps, data services, and machine templates. From a usability perspective, the most valuable aspects are its custom activities, libraries, and object repositories. In terms of integration, I like the ability to use APIs and call automations from UiPath apps. The most valuable feature from a human-in-the-loop perspective is the action center. 

Our customers appreciate the support that UiPath provides, and they don't want to go with a third-party vendor like Microsoft Visio, Form Recognizer, or Google Cloud. They're hesitant because some integration is required. The lead times for closing queries are longer with third-party vendors. For instance, it takes me about two or three weeks to set up Document Understanding in my project. But it took us three months to establish Form Recognizer with a client.  

In addition to the out-of-the-box functionality UiPath provides, it can host our custom models. That's something that comes in handy when we need a custom model. So far, we haven't taken it to production yet, but we are still baselining the technology. At the moment, we are doing a baseline project where we try to perform four POCs simultaneously. We are baselining Google Cloud Platform, Azure, and AWS with UiPath's AI center and machine learning services and comparing the four.

UiPath's performance could be improved. UiPath's framework was built on top of .NET Core. It was a 32-bit platform initially, but they recently introduced a 64-bit version. Let's say I have a huge machine with 64 gigs of RAM. If I have a server machine and want to use multi-threading to extend my automation and multitask, the design won't allow me. I can't separate things into multiple processes. 

The platform is designed to go step by step. Parallel activities are not truly parallel, but it creates the impression that it's running in parallel. For example, if you're on the left segment within a parallel activity, and there is some wait time, it doesn't stay there. It goes to the middle and then to the right. It schedules tasks based on a time-to-completion window and then takes them from end to end. 

UiPath optimizes the time and doesn't let the CPU idle, but it doesn't give you multi-threading or asynchronous methodologies. These are available in the C# and .NET framework but absent in this platform. It's a step-by-step process where you go through each activity. A casual developer or coder who wants to leverage UiPath should be able to. I'm not saying that the working code is not there, but it's quite basic. It doesn't support functions or asynchronous methodology. 

UiPath is attempting to make it easier for a citizen developer to automate processes. They don't have to know how to code, but a citizen developer can't do it when the use case becomes more complex. When they advertise that one doesn’t need to know coding to program bots, that's only true for easy or intermediate use cases. We still need a programmer for anything beyond medium complexity.

The marketing could be improved because the methodologies went from waterfall COE to an automated operation model. However, people are trying to do automation in an Agile model, but it's not exactly executable that way. When customers see the demos from UiPath, they expect that the results will be significant, and they are. However, we might try to automate something, and we’re unsure whether it can be automated because there's a gray area. There's always a 20 to 30 percent chance automation might fail. And that gray area is something that I want UI to focus on.

They have tried this with StudioX by adding checklists. The industry is not following this practice, though. I'm not sure how they should ensure that it gets followed within the platform, but the delivery model needs to improve. It's still niche. 

Another thing to consider is the work-life balance of the developer and the solution architect. The overall challenge of automation tends to become exponentially complex over time. For example, let's look at one aspect: the account tables. I can go to the account tables from a simple PDF perspective. The PDF is readable by the board, and the solution can extract all the data and do the account tables within SAP or Ariba and mix all of it and then submit a report to the business.

This can be extended to intelligent document processing using form recognizer and custom models, then passers, pre-processing, post-processing, and sending the report to the business. The complexity of it can be extended quite a lot. There should be a framework or methodology in place to hedge the bet so that it's not too complex and doesn't disrupt the life of a developer, solution architect, or business analyst. 

If the automation becomes too complex and challenging, our support team won't be able to sustain it in the long run. Once the development team is gone, the automation will die two or three months down the line. It's a balance to manage the complexity and extent of our automation.  

We've been using UiPath for a little more than four years. 

Resource utilization is one area where UiPath is lacking. UiPath says that the solution will run fine on a machine with four gigs of RAM, and they recommend horizontal scaling, but I suggest a mix of horizontal and vertical scaling. 

I've seen implementations on giant machines with high-density VMs and five users logged into the same VM. Therefore, the resource utilization isn't optimal. The RAM and CPU are not completely utilized. It only executes processes on a segment of the resources. I think that can improve. 

I rate UiPath's support nine out of ten. UiPath's support is excellent. They triage issues based on severity, and there is a clearly defined close time and lead time. Their support engineers will follow up with you 24/7 over phone, SMS, or email. 

The scope of support isn't limited to problems with the UiPath platform. We can reach out to UiPath if we are having problems automating a third-party application. They will help us if they have experience with the app. If they don't have experience, they baseline the issue and go through the log to do whatever they can to help us. We've had a great experience with UiPath's support, and our clients feel the same. Support is one reason UiPath is dominating the market. 

Positive

Most of the RPA solutions are fairly similar. The inspiration for UiPath's object repository was taken from Blue Prism. UiPath's integration services are like the connectors in Microsoft Power Platform. I'm not saying that UiPath is exactly copying everybody, but they're taking the best features from every solution and bringing them in-house. 

Other platforms are dominating in some areas. For example, Power Apps is more mature than UiPath Apps. I'm trying to add value based on my experience, and Power Platform's connectors should also bring value to UiPath. In the end, it shouldn't be redundant.

Every time we deploy the solution, we use an automation operation model. It's a massive document with policies defined on every level, from design to development, UATS, prods, escalations, business, teams, team leads, Agile boards, and reporting. 

All of that is documented from the start. We use that model to layout deliverables needing to be fulfilled. Once deployment progresses from one step to another, we have a way to document our progress. We've gone from a theoretical model to a UI model. It's not purely Agile or KanBan, though Agile framework and KanBan breakdown structures are there. However, it doesn't follow a scrum methodology. 

We're not on a two or three-week release cycle. One sprint is the entire use case from build to development and then from development to UAT to production. It's a custom delivery model, and it's working. Still, I feel it can be improved. 

Our clients have seen significant returns using UiPath, but their marketing could be improved. 

I'm aware of how UiPath's pricing compares to other tools, but it's hard because the offerings are different. It's not apples and oranges per se, but it's comparing an average tool to an excellent one. UiPath provides enormous value, so the licensing is justified.

I rate UiPath nine out of ten. It isn't perfect, but they constantly improve and surprise me. At the moment, I give it a nine, but it might be eight in the future. If you feel like some process will cause a lot of headaches, position it later in the cycle of automation. If you can save resources by automating, you should go for it, but you should be smart when deciding your use cases.

If you're thinking about implementing UiPath, I recommend having a design team that understands automation. You need people with some experience who know how automation is done. It requires some business analysts with at least a month of experience on UiPath from a citizen developer perspective. It would help quite a lot in terms of establishing automations that are relatively complex. Try an 80-20 approach operating principle when planning your automation.

We primarily use the solution for the typical banking and transactional use cases. An example could be something such as homeowner's association transactions. That incorporates multiple departments in a bank, including treasury management accounts and opening an ACH. All of those areas are very transactional in terms of the rules-based processes that they follow.

We also use the solution for wire transfers. We have a lot of those. Then, from a business perspective, we look at lines of business as well. It can be used for residential and commercial lending. It's very process-driven and very transactional. We're able to incorporate automation in those areas very easily.

The solution allows us to do more with less. If we have a relationship with a third-party bank to onboard with a thousand accounts with a 15-day turnaround time, depending on what the relationship is, without automation, without RPA, you'd have to hire at least 10 or 15 contractors to get that done. That’s from a contractual perspective, to make sure that you meet the deadlines. However, automation just allows you to scale. You can save the same scenario in multiple areas. The scalability and the time to scalability on offer are key. And that's where we see the value as an organization.

The time to value is excellent on this solution. It is incredible. Unlike any other technology, which is really generally workflow-related, this RPA will give you quick wins, as demonstrated by the return on investment.

The solution has saved costs for our organization. We look at it from different lenses. We look at it from an actual savings perspective. We forecast it and then we annualize it to understand where the product will be or what it will bring us in the long and not just short term. For us, it’s very ROI-driven.

I’ve been with the bank for nine months now, and we went from having two processes automated to now having actual savings of about a little over $300K. It's not bad. There are about 30 processes in production. If we annualize that and forecast out we can see how we can continue to grow.

We have been able to reduce human error. I would call that the cherry on top. You don't really track that until you have it automated. However, if the alternative is having end-users upset that mistakes happen, having the level reduction we’re getting is great.

The solution has freed up employee time by a lot. We see it wherever it's very transactional heavy, where we had multiple team members. Now the teams are built by bots and people. Where you had three people before, now you have a person and a bot executing. So far, we have likely saved, in these nine months, a little over 11,000 man-hours. That's quite a win.

This additional time enabled employees to focus on higher-value work. That’s the whole purpose. We tend to look at ROI, not just money. It's also what else are the employees doing. We’re saving money from a man hours perspective, and, from a cost avoidance perspective, we’re not having to hire. The folks that work for us actually can do valuable work. They can focus on better decision making, more time with the customer, better relationship making, et cetera.

My team has started to utilize the UiPath Academy. My team is composed of multiple different skill sets from PMs to BAs to engineers, and so forth. Even at the level of IT, with the teams that manage infrastructure, sometimes I have them taking some of the infrastructure courses in the Academy. The Academy works well due to the fact that they prep you to grow quickly. If you don't understand how the technology operates, it could be a little bit difficult. You do have to train a little bit to understand it a little bit, however, it's not very difficult, especially when you have access to the Academy.

Having the Academy has affected the process of getting employees up to speed big time. Their appetite to get better and better is just enormous - not just due to the fact that the product is good and it's providing that information to do that, but also due to the fact that they're seeing results. That's helping them want to do it better and faster. It’s a win-win situation there.

The Academy has a really good curriculum for very particular skill sets. It's hard to be an engineer and have to learn a bunch of different stuff that won't necessarily concern you. UiPath has been able to really compartmentalize that learning capability so that it makes sense to different groups and speaks to their level of expertise. That’s probably the most valuable aspect of it.

In the world of NextGen technology, it would be ideal if they could speed up their beta products. My team handles all next-generation technologies that we implement in the bank. For example, we're looking at task mining right now. It's theoretical, however, that product is going to be tremendous. That said, it's still a bit in beta. For us, if they were trying to get new items a little bit quicker out there, that would be ideal. They're running with it, therefore, I'm not overly worried. They also want to provide a good platform for its users, and so I understand why they want to make sure it's good. The competition, in reality, even though it's growing, is not as strong. The feedback is just me being picky at this point.

We've been using the solution since 2015.

So far, we have had very few issues and the issues really have been around very complex knowledge. We have reached out to UiPath to get advice, and they responded, and it hasn't been bad at all.

Scalability has not been an issue. It's not even a topic of concern. The concern is more on the business side and can you manage the growth from a business perspective. The technology's there and I want to automate more. However, with automation, the question is if you can manage it.

We do have plans to increase usage. 

We have about 30 automations in place. That means about 60 to 70 people are using it. They work in shifts and they'll come in and they'll use their bots to process stuff while they do other stuff. 

Technical support is great. In the times we've used it, they've come back relatively quickly.

I started with UiPath a long time ago and I actually worked for Automation Anywhere. I was head of their cognitive product. Therefore, I used them for a while. They're not a bad product at all. They're really good, however, their scalability is a little bit of concern. For us, coming back to UiPath was probably a good decision.

The initial setup is complex. However, that's okay. It's also not supposed to be easy. You have to learn a little bit about what you're doing. Any technology would be a little bit complex. Once it's in, you can get that learning curve moving. With the actual users, that's where the trick of the implementation piece comes in. Once you're through it, however, that's where UiPath becomes really tremendous.

The deployment was maybe 60 days, however, it wasn't necessarily just because of UiPath. In banking, there's red tape. It takes time to get certain aspects approved. It takes time to get the right infrastructure in place. The enterprise has to be willing to put the time in just as UiPath has to put a certain time. It's a blend. As we were implementing, we were still putting things together and setting up governance. It was not just the technology. You still have to prep on the business side.

In terms of maintenance, on my team, I have about seven people, three engineers, a couple of business analysts, and a couple of PMs.

The pricing could be a little bit better for document understanding. I ran the pricing scheme for intelligent documents. With understanding at Automation Anywhere, there's room to maneuver. They can do a little bit better in making it easier for customers to dive in. I know it's available, however, it is what it is.

The organization evaluated Blue Prism, Automation Anywhere, and UiPath. 

We went with UiPath due to the fact that they gave us a good deal and we wanted to try it out. We started little. At the hedge fund, we're actually now at over three bots in production. Automation Anywhere's a little bit pricey at the moment. We just kind of pulled the trigger on UiPath and it was a blessing. 

We're just customers and end-users.

While we are on-premises, we're starting to use the cloud including some of the modules from the cloud, document understanding, task minding, et cetera. We're using the latest version of the solution. 

I'd advise those considering the solution that there's a secret sauce to it. It's not just the technology. People need process systems and a really good partner. That's how you get it done.

I'd rate the solution at an eight out of ten.

We have robots and we have Orchestrator and we are exploring the new analytics model over the next few months. Right now, our primary use is mostly operational processes and deploying apps for the global operations team. We have multiple automation processes in place for them already.

The product has improved the way our organization functions in a lot of ways. Mostly, I would say it improves our processing efficiency. When you have one person working eight to five to take care of tasks and can replace that, instead, with a bot that can work 24 hours a day to onboard customers or entering information to a form, that drives a lot of value for the organization. With that benefit and decreasing human error as well, that turns into a lot of value for a global organization like us.

The most valuable feature in UiPath is different from one process that we build to the next. I would say that the ease of configuring new processes may be one of the most consistently valuable features for our robot production. Creating a process is just a drag-and-drop solution most of the time. With this automated feature for process creation, it is very easy to make what we need, and that ease-of-use is valuable.

There are a few areas where the product can be improved for our use. Invoice processing, for us, is a major use case. I saw some examples in our research that mentioned machine learning models and how to implement that in the new, upcoming version. It seems that the machine learning would solve some of our current issues in processing. That is one feature that I would like to see and experiment with when it is released. I want to explore that to see not only what it can do, but how efficient it is and how it affects the performance of invoice processing models. The other feature we are interested in that is promised to be in the next release is the VDI (Virtual Desktop Infrastructure) desktop. 

Those two features, for us, are critical. But we also need them to be fast and efficient for us to be able to use it in practice. Otherwise, we are going to continue struggling with the solutions we already have. The solutions are kind of already there in UiPath, and I think they need to work hard on the performance of anything they release. This is especially true for those two solutions. For us to have a good business case for deploying them for use, we need them to be efficient.

A feature that we tried to use that clearly has room for improvement is the UiPath Computer Vision component. We have tried to use that but it was not efficient enough for our planned applications. It was a little disappointing. It needs to be improved in performance and in the design of the machine learning models. It is pretty easy to use for people considering the technology. I just don't think it is quite to the point where it should be.

On a scale from one to five with one being the least stable and five being the most, I would rate the stability of UiPath as pretty stable. For us, it has really been a five.

We have 35 employees involved in the automation program doing configuration. We also have three platform administrators and a bunch of automation champions right now — there are quite a few of the latter.

Customer support, in general, is pretty good. They are very responsive. We have one dedicated support person in New York and we always get support right away. But there are other facets to UiPath technical support.

The classroom and Academy training are both easy to use and beneficial. On a scale of one to five where five is the best, I would rate these resources as a five-out-of-five. It was and is beneficial and it is available to help the staff get oriented to the product and resolve production issues.

We have 35 configurators. Of those 35, I will say that only five or ten of them were formally and personally trained by a trainer at our site in Costa Rica. The rest — and really most of the configurators — have been trained only using the UIPath Academy suite. So, with some senior resources available and the Academy, you can establish a team of proficient configurators in a very short time.

The reason why we choose UiPath was because they let us play with the tool before we bought into it. We were looking at many vendors for an RPA solution, but the other vendors wanted us to pay first before they would let us explore the products and what they could do. With UiPath we were able to play first without paying and that's something that was attractive and showed that the company was forward-thinking and confident. We explored the product and saw that it was something that could help us to solve some of our problems. So that is why we identified it as a solution and eventually selected it as the product that was going to cover our processing needs.

The initial setup was pretty easy because we had the assistance of UiPath experts on-site. They helped us with the first deployment. They did a lot to help make it pretty easy. It might be more accurate to say it was not easy, but that because of them it went smoothly. That's what their group was there to do: sort out modeling problems pretty quickly and get the product to production.

We used a consultant and our experience with them on a scale from one to five where five is the best, I would say they were a five-out-of-five. They knew what they were doing and were excellent in promoting the deployment. We have no complaints about their services or the result.

We have automated a lot of things and realized about a million-dollar return on investment in about a year-and-a-half. There is still a lot more to do in our company. We are big, so we have a lot of opportunities for automation and we are expected to be producing even more benefits than we already have. But a million-dollar return on investment is pretty good and is only the beginning of what we will eventually realize.

Right now we have around 200 licenses in terms of the bots we have in production. I know the cost is about $200,000 or $300,000 per year. I am not sure about additional costs.

From what I understand UiPath is cheap when you compare the costs to other vendors in the same market. In any case, the price seems to be good in comparison to the actual benefit. We also have some type of special agreement for pricing discounts because we are one of the early adopters who engaged with UiPath for use of the product. I feel we have some very special treatment and the price, for us, is reasonable and convenient.

On a scale from one to ten where one is the worst and ten is the best, I would rate this product as an eight out of ten. The reason why it is an eight and not a ten is because there are some specific considerations — especially in performance and machine learning — that we believe has a lot of room for improvement. They are starting to introduce the functionality, they are doing a good thing in introducing it, but there could be a lot of improvement.

If I would make a recommendation for people considering automation options, I would say that they should take advantage of manufacturers that let you play with their product to evaluate if a particular solution is convenient for you. This is the reason why UiPath became the more convenient option for our company when we were looking to start with process automation.

For us, the virtual implementation has been working well because we have deployed everything that we automated in our Citrix environment. Now the new question is how easy it will be to interact with target applications through our VDI desktop. Resolving that to this point has been pretty tough for us and it is actually one of our constraints in making processes work efficiently.

Robotic processing has helped to eliminate human errors and reducing human error is definitely one of the basic benefits that those moving into robotics should expect — if they are approaching automation correctly.

The solution also saved our organization time. For example, we had the one case where we had to onboard around 2 million customers. What that would take in terms of manual hours is about a month and a half. Instead, using RPAs, we were able to complete the task in one week. That is just one example. We have multiple examples in three years of automation. 

We use Studio, Orchestrator, attended and unattended robots.

We use this solution to solve the things that people don't want to do. They spend a lot of time and there is a high potential to make manual errors. Quality suffers because it takes too long and users can get fatigued. There are the things that we are targeting and we have already seen some of the benefits.

We do not run our automations in a virtual environment.

With respect to how easy it is to automate our company's processes, on a scale of one to five, I would rate this solution a four. It is easy, but there are some improvements that can be made.

On a scale of one to five, judging how beneficial it is, I would rate the training a five. It is my lifeline. 

From the point that we purchased our UiPath license until we had our first robot was approximately six to eight months. I felt that the process was complicated, in part because of UiPath but also partly because of internal things.

Internally, in order to get set up, the entire infrastructure needs to be in place. All of the servers have to be set up and you need the right permissions because the bots need their own security. You have to explain to people that this is going to be a service account, and you have to explain the need for it. These are all internal, but necessary issues.

This solution has saved our organization time. I would estimate that the bots save us four weeks every month.

In terms of eliminating human errors, this solution absolutely helps in that regard. The bots don't make any mistakes. Rather, they point out the mistakes that have been made by humans so far.

The ability that this solution has to work with Excel is really good.

The SAP feature looks helpful and we are going to explore that functionality in our next project. 

We don't like the attended bot experience, where it requires the user to be hands-off. We installed it on the user's machine, and the user should not touch the mouse or keyboard because the moment they do, it crashes. It takes over the machine and cannot work in the background. For this reason, it is difficult to work out scenarios that are purely attended. 

For example, we have a process that downloads two files and then does some massaging of the data. Now, the data needs to be looked at by the user. An email is sent and the files are deposited into a common folder, and the user will address the task when they have time. After the user manipulates some of the data or makes decisions, it can proceed to the next step.

They drop the modified files into the common folder, but the bot doesn't react because there is no feature in Orchestrator to trigger based on the arrival of a new file. Instead, we have to put in a call to check periodically, whether it be five minutes, fifteen minutes or some other interval. This is something that we don't find desirable.

Creating a trigger that is based on either the arrival of a file or the arrival of an email is a common scenario and it should be built into the platform so that I don't have to code for it every time. As it is now, there are only two ways to trigger a bot.

Orchestrator does not have a good filtering mechanism to look for jobs, and the table view needs to be improved.

From the UiPath side, one thing that irritates me is that you cannot find the downloads to upgrade to the new version. I'm a paid customer and I log in with my credentials, but I cannot download it. I need to create a ticket, tell them who I am, and I have to give them a long list of things that I don't remember. They will address the ticket maybe today or tomorrow, and then finally I get a response. When I use other tools like Microsoft and SAP, as a customer or subscriber of the tool, I just download the latest versions and install them on my servers. When I have a mandate to do work, I want to be able to do it. However, when I can't download the file and have to wait for a ticket to be answered, this wastes my time.

The integration with Elasticsearch and Kibana is a struggle. They are not UiPath products, but they are recommended by UiPath. Some documentation is provided, but it is an Orchestrator installer package the just installs itself. I had to do a lot of experimentation on Windows machines because the configuration is different for Linux machines, and this gave me some trouble. There is a lot of information about this in the UiPath forums and I spent a lot of time on it.

Right now, I have a set of configuration data that I put into an Excel file. The users can change the file and my app will run using the configuration variables. A problem occurs if the Excel file is not closed properly because there are locking issues. This is a pain because the bot crashes when opening a locked file. There is something in Orchestrator called Assets, but it is kind of limited. It only accepts text and a credential. It would be nice if it had a simple database table, say to be able to create a couple of columns that I want to set up for a bot-related task. Rather than a single piece of text, there is a whole table of information. My users will be able to edit it, based on the permission that I assign. Then the bot will look at it and work according to what is specified. This would be a great feature to have.

I would like to see the Studio web-based so that we don't have to install it on everybody's desktop. 

With respect to the stability, on a scale from one to five, I would rate this solution a five. I have never had any problems with the platform itself. I haven't had to restart the servers or anything like that.

We have three people in the organization who work with the bots.

We did not use another RPA solution prior to this one.

In deciding whether and RPA would benefit us, we followed the guidelines that UiPath provided, which is a twelve-step validation process. If processes are unchanging, there is a lot of manual work, there is the potential for errors, and it's simple to automate, etc, then RPA may be suitable. Based on this assessment, we made the investment in RPA.

Our own team was responsible for the implementation.

We have recognized ROI through savings in employee hours. For our PoC, with two projects, we saved one full-time employee.

It took us one year to see ROI, although I don't think that we have saved any money yet because of the development time. If we negate the development cost then it might be $60,000 USD, as a developer's salary, but we're not there yet. One of the things that I am trying to do is motivate people to look at more processes that we can automate because right now, my bots are sleeping and I want to put them to work.

Our licensing fees are approximately $140,000 USD annually, which includes all of the bots, Orchestrator, and the Studio.

In our company, there is another group that had already implemented UiPath, and they have had good success. We followed their lead, but in addition, we reviewed other sources such as the Gartner ranking. The features set was also considered during the evaluation.

I feel that the cost of the bot is worth it, provided that we make use of it. The unattended bot is cheaper, but it is useless for us right now because there is no use case. We think that Studio X might change that, and I've heard that the Studio X license includes the attended bot. So, if we swap the attended bot license that we currently have then we could take them away and then get the Studio X license, and that will motivate more people to make their own automations.

No organization has unlimited resources, but the business is changing around us and we are always tasked with new things to do. In that aspect, you have to make room for innovation, and you have to automate.

UiPath has shown tremendous gains with this solution. They're sitting on the shoulders of Microsoft .NET, and they've shown some initiative on what you can do with a generic platform. They offer free training and a Community Edition for people to experiment with, and it can do wonders for the world. We have seen that happening and I love that.

This is a company that is listening to customers' feedback and I think that they should keep doing that.

My advice for anybody who is considering this solution is to start by watching all of the videos. Go through the UiPath Academy and get a feel of what it can do. Read all of the case studies and see what other people have done. You will get a feeling for the ROI. Then download the Community Edition and play with it to see for yourself what gain you can get from this tool. Finally, start small and just keep adding to it.

I would rate this solution an eight out of ten.