What is our primary use case?
Our use cases are a lot of both data scraping and data reconciliation. We tend to find that we're going into web portals, we're going into spaces that require a user interface, and extracting some kind of information or data from those places. Then, we transform them. Sometimes it's just as simple as saving it as a different type of data. Still, it can sometimes be rather complex to combine that data with SQL queries to enhance it, or even just clean up and remove extraneous data that our users don't need.
We touch on other areas, such as averting risk by double-checking things or replacing manual transcription. We find that due to the size of our team, we're relatively small, and the biggest bang for our buck tends to be in the data reconciliation and gathering.
How has it helped my organization?
We have about 40 bots in production right now, which is pretty good for a small team. We're only really two developers right now. We're able to get pretty good value out of those bots. They're running probably about 50% of the time, which is pretty good since, as a financial organization, we're really seeing peak hours in the early morning and late evening as we're getting in all of the morning data and closing out for the day. We have a pretty good amount of utilization when we're busiest.
What is most valuable?
It is easy to build automation. For building them, we find that there are two categories. There are some that are quite simple and some that are quite quick, and then there are some that require some additional thought, some additional background knowledge, and some additional expertise. One of the things that we like about UiPath is it really does have a lot of pretty solid integrations. They can really get us what we need relatively quickly, but we still find that there's a level of expertise required to make sure that everything matches the way we expected it to, whether that's formatting or the data types. We need a little bit of expertise to fit them all together. However, the pieces of the puzzle are all there. It's just a matter of finding where they go.
We use a lot of unattended robots. We're almost entirely unattended right now. We find that those are the best early bang for our buck. We were able to completely replace a manual step from beginning to end. That's where we focused our efforts.
We're interested in expanding into attended; however, we're still looking for the best way to find the more business user level, or the champions who need some knowledge of RPA to be able to utilize it properly yet also have the flexibility to be doing their normal day job and their normal day-to-day tasks without it interfering.
We have about 40 automations in production. Those are business related. That said, we are also starting to add a few more now. Probably on the order of five at the moment. We're hoping to continue to expand it in what we refer to as maintenance processes, things that allow us to automatically restart our own jobs when certain restart error conditions are met, and things that allow us to monitor the health of processes and make sure that there aren't missed exceptions over the course of our day-to-day. Therefore, we have about 45 total, with 40 business related, and then we're slowly growing this number that allows us to better and more efficiently perform our automation functions.
Those five robots are dedicated to maintenance handling exceptions and things like that. One of the things they do, for example, is scraping the logs and they'll look for keywords in our logs. The obvious one is exception or error; when they see those, they'll flag them and bring them to our attention.
We're currently working on expanding that functionality so they can be a little bit better about proactively helping us. We're currently testing the use of the API calls within those monitoring jobs to restart ones that hopefully, just a simple restart will fix. For example, a slow or broken web portal that otherwise wouldn't be an issue.
That's a big time-saver. Our bottleneck as a small team has been how we support them as we get more and more processes into automation. We decided that the best way to help ourselves is to automate some of these support and repair tasks. We see that as being a big boon going forward to us.
The most valuable feature is the flexibility it offers. It really can integrate with just about anything you need it to. It has places where it is preferable to be used; however, in a pinch, you really can finagle whatever kind of circumstances you need out of it. We found that a lot of the time, it's not the perfect tool for the job, yet, thanks to the speed and agility with which you can apply it, even if it's not going to be a solution for 12 months, for six months it might be exactly what you need to get somewhat over the hump.
The user community has honestly been really impressive to me. The UiPath forums are genuinely a spectacular resource. It's really rare to come across an error, issue, or challenge that hasn't been at least discussed to some extent on the forums. That has been a great value add for us. Even if something is eventually going to be technically possible, you can frequently at least get an idea of how difficult it will be to implement it by seeing how much people have talked about how hard or easy it is. It really gives you some flexibility to quickly assess how doable it is, and then make a decision to move forward with it or drop it.
I used UiPath Academy several years ago now, in 2019. I know it's pretty different now. My team members have also used it. The courses are very valuable. There's a lot of great knowledge in them.
What needs improvement?
One of the improvements that could be made is the support interface. I've talked to one of the product experts today about how they could better show when there are issues with a particular job. Right now, we tend to find that everything just looks okay unless we really go out of our way to highlight issues. What we would really love is for there to be a way to flag in Orchestrator exactly which jobs had warnings, exactly which jobs had business exceptions, exactly which jobs had exceptions that were handled, and which ones had unhandled exceptions. Right now, we are forced into this success or failure state. The lack of granularity makes it very difficult for us to quickly and easily find where we need to address issues.
There is no out-of-the-box way to configure granularly. This is where our maintenance processes are starting to come in. They're trying to fill those gaps, yet realistically, it would really be better to just have those visuals right there in Orchestrator without us having to rely on making our own API calls and checking for special criteria outside of what UiPath would be able to detect on its own since we're setting it within the code itself.
The UiPath Academy courses are disorganized. We have trouble identifying exactly what we want people to learn from within the list of courses, and since they've changed so much, it's sometimes hard for me to know what my coworkers are learning versus what I learned. I don't know where either my gaps or their gaps are. I don't know if I need to assign extra training or if maybe I need to go back and take a newly added training. That can be difficult for us.
For how long have I used the solution?
I've been using UiPath for about four years now.
What do I think about the stability of the solution?
The stability has been good. I don't have any complaints generally about the stability. There aren't any particular issues I could name that would cause me to have a negative opinion, which I would say is probably about as good of praise as that perspective as you can get.
What do I think about the scalability of the solution?
We are not using the cloud. I know that the scalability is a lot better with the cloud features. We are an on-prem using preassigned VMs. We really love the ability to bring them on and off as needed. However, I will say that I know there's a whole universe of adaptability out there that, unfortunately, I am not allowed to partake in.
How are customer service and support?
I have worked with technical support a little bit. I would say they are decent to good. They have been timely responses. Still, they are not always as detailed or helpful as I would like.
How would you rate customer service and support?
Which solution did I use previously and why did I switch?
I did work on Blue Prism for about two years.
It's been a few years now, so I can't speak to the current state of Blue Prism. That said, one of the ways I really struggled with it, was it just did not have the connectivity that UiPath has out of the box. A lot of applications we use regularly, such as PDFs and Excel documents, just did not work intuitively. I remember specifically having to write the code to handle passworded PDF documents manually. And so, that lack of integration in things that you typically use in every single process was a real frustration with Blue Prism.
That was one of the first things I noticed about UiPath is they handled those integrations much more smoothly. There are many more of them, and they are much more ready. The other thing that I would really give UiPath credit for is they are much quicker to iterate and build out new helpful features. Blue Prism was frustratingly slow to fix these problems once identified. You would face the problem and be forced to solve it with your own code since you couldn't rely on Blue Prism to get a solution to you in a good amount of time, unfortunately.
How was the initial setup?
I'm not involved in the infrastructure deployment.
In terms of robot deployment, it's relatively straightforward. UiPath offers good tools to allow us to do it. My company, in particular, needs to streamline our ticketing system and our release process a little bit. Our particular process makes the best use of the tools available due to internal limitations, which we're planning on working on. For example, we currently do not integrate with Git, and that's something we need to fix.
What was our ROI?
My understanding is that we have seen a return on the investment. From my perspective, we just need to continue to highlight the value add and make sure that people are aware of what we're able to do to further expand the automation. There's a lot of untapped potential there.
What's my experience with pricing, setup cost, and licensing?
I don't get too much into the licensing and costs in my role, so I can't really speak to them.
What other advice do I have?
We do not really use AI/ML capabilities. We're a smaller group; we just don't have the resources to learn and utilize those tools the way we could. We're in the process of moving over to some better workflow management and BPM-style tools. We're trying to build one in-house. That will allow us to take advantage of some of those machine learning and data integration features, and then once those are fleshed out, and our business teams can continue to get onboarded into the BPM workflow, we'll be able to start utilizing those tools more heavily.
My advice to others would be to go through some of the UiPath Academy pieces of training to see if any topics are covering issues they are facing. They generally do a good job of showing you how to tackle those problems. Do a proof of concept and validation of it. UiPath is a tool. It's very good at some things, yet it's not very good at everything. Like any other tool out there, it's an excellent product if you find that UiPath is the right tool for your job. That said, you don't want to shoehorn it into something that really could be done better elsewhere.
I'd rate the product a nine out of ten. UiPath does a great job of iterating on its product, adding new features, and combating the negative sides. I've really been impressed by how quickly they do that. Every time I have a criticism, I find out that somewhere in the pipeline is a new solution coming to help me with it. That said, they're not always perfect when they come out. Sometimes we don't adopt them until after they've been iterated upon a couple of times, yet, just the fact that they listen and build these solutions is really helpful.
Disclosure: I am a real user, and this review is based on my own experience and opinions.