UiPath Process Mining is as lightweight a tool as it is new. It's not able to handle a lot of logs in one go at high speed. The solution's speed is fairly good, but it should be further increased to make it more powerful to process large amounts of logs. Logs should be as much as possible. We have taken one year's and sometimes two years' logs to understand the behavior.

We are not here solving small problems because a lot of money is invested. I have observed that handling a large amount of logs slows down the solution's performance. This area is improving because we are now observing a new version of UiPath Process Mining with improved speed. As a developer, I'll be happy to see that happening more quickly, and I am sure the solution is heading there.

UiPath Process Mining has a cloud-based version and doesn't have a desktop version like Celonis. Celonis has a desktop version, which is very easy to use. The latest version of UiPath Process Mining has recently integrated task mining. However, Celonis has something called the Execution Management System, which is much more advanced and powerful.

Some of those capabilities can be built into UiPath Process Mining. Data integration should be made easier in UiPath Process Mining.

It would be beneficial to improve the cost since it is quite expensive.

The inability of Process Mining to address bottlenecks across a variety of processes suggests potential for improvement. While the tool provides details at the bottom of the page that can be adjusted for more or less granularity, there is room for enhancement. It would be beneficial if the program incorporated AI suggestions, such as recommending decreases in certain approval points. By leveraging advanced AI algorithms, Process Mining tools could potentially offer proactive insights and recommendations for optimizing processes, contributing to further improvement in their capabilities.

While exploring other Process Mining tools, I found that most of them effectively handle processes like Purchase to Pay and Sales Order, but they lack features related to manufacturing or production processes. Specifically, there is a need for connectors that can integrate with production parts. This gap represents a valuable opportunity for improvement in these tools.

Improvements could be made in educating a new support team for enhanced efficiency and effectiveness.

UiPath has an in-built feature that calculates rework. In most cases, this calculation by UiPath is wrong. UiPath intelligence does not consider that the definition of rework varies for every process and firm. It just considers the normal definition that if a step is repeated twice, it will be a rework. Sometimes, repetition might be necessary for the process.

When we understood the issue was with UiPath intelligence, we deleted it and created our own single-value KPI for rework. The intelligence KPIs offered by the solution are not up to the mark. I don't know if this feature was changed. I am not sure about the current situation.

Streamlining processes using UiPath Process Mining can be challenging, as it requires initial clarity on various aspects. We need to understand the process scope upfront, including whether UiPath has the automation capabilities to address it effectively. With this transparency, UiPath experts can provide guidance, facilitating a more efficient process mining experience.

We encountered difficulties when trying to integrate UiPath Process Mining with SAP. I would like to have greater integration capabilities.

I would like UiPath Process Mining to utilize more AI in the process.

UiPath Process Mining should focus more on developing connectors to source systems. In fact, I worked on an interesting consulting project for UiPath Process Mining when I was in another service industry. This was when they acquired the process mining tool Process Code in 2019 and rebranded it as Way Beyond Process Mining. They came to us because they wanted to compare their tool to the top five process mining tools on the market. We conducted a consolidation project and presented our findings on the strengths and weaknesses of UiPath Process Mining. UiPath Process Mining has come a long way since then, but I believe there is still room for significant improvement.

UiPath Process Mining has not been adding new API connectors as frequently in recent years. As of now, there are only around ten to twelve source connectors available. In contrast, other market tools, such as Celonis, offer much more extensive connector libraries. This means that UiPath Process Mining can only connect to a limited number of systems, even though it can connect to standard systems like SAP and Salesforce. If we need to connect to a new or less common system, we will need to develop a connector from scratch, which can be time-consuming.

UiPath's UI and assets could be improved. Celonis performs better in terms of UI, even though it uses fewer colors. However, UiPath's process maps are congested. While I understand that organizational reasons may require complex processes, I believe that Celonis provides better visibility in this regard. Therefore, I recommend that UiPath improve its process maps along these lines.

The number of file formats that can be uploaded should be increased. For example, the BPMN file format is very important, but it is not currently supported by UiPath Process Mining. 

Multiple users with different access levels need to use UiPath Process Mining. For example, if I develop a UiPath Process Mining dashboard for the leadership team, they need a certain level of access that allows them to view the dashboard without seeing all of the details. I also need to provide access to end users who are auto managers, but who may not be technically proficient enough to navigate the tool. Therefore, the current number and types of user profiles available in UiPath are not sufficient. We need more flexibility in providing user profiles to meet the needs of different users.

I'm not involved in finance but I know this is an expensive solution. It would be helpful if licensing costs were reduced. 

Process mining is something that we are doing, but I think it needs to be more mature. UiPath is still new to process mining. They are decent in SAP ERP when it comes to process mining. But compared to Oracle's E-Business Suite and all, they are still pretty new. And they need to come up with some model templates for how to adopt process mining within your organization. For example, they could use some accelerators to help onboard customers. Right now, it's not that easy. So process mining is something that they need to improve on for Oracle ERP.

UiPath has an excellent strategic roadmap when it comes to features. So if it's not already in the works, one thing they could add is AI fabric, and if UiPath could bundle it into the primary licensing with Orchestrator, that would be great. Concerning artificial intelligence and NLP machine learning, we are getting rid of the requirement where a chatbot can integrate and synchronously interact with the robot in real-time.

The ability to listen to and reply things in real-time would be good to have. They are already doing it, but the ease of implementation and options to integrates needs to be flexible, and the license should be bundled into their current offering so that it's more cost-effective.

Another thing is UiPath's OCR we hope it has the same quality as FlexiCapture, a separate solution by ABBYY. That would be awesome if the UiPath had native OCR functions that were the same level of maturity as FlexiCapture.

Within UiPath, AI Fabric is a valuable feature, but Process Mining can be further improved with AI. AI is booming. Once they improve the AI capabilities, more and more customers will come to UiPath. There are a lot of tools with AI capabilities.

The installation documentation and guides for users from UiPath Process Mining could be a little better.

We faced issues with the data sharding process. UiPath has its own limitations in terms of the number of rules of data it can actually handle at a given point of time. As a tactical step, they provided the data sharding approach, but that was not the best. It kind of complicated the analysis rather than making it easy for people to analyze large data sets. So, the handling of large data sets is a major problem we want them to improve.

I think they can improve their overall feature set around the visualizations, which they can provide out-of-the-box. Some of these visualizations are not already there. 

Finally, it is about the ETL process. I think the ETL process could have been much simpler compared to what we do right now, which consumes a lot of time. If I talk about a typical project life cycle, I would spend about 30% of the time getting the data into the system, which is obviously not ideal compared to some of the other tools we have in the market.

Their ability to create out-of-box integrations and solve for the ETL problems could be an improvement, and their ability to help us scale and do parallel processing when handling large data sets. They could also improve their overall look and feel on how the analytics will show up on the tool. They just do the job, but when I'm showing an analysis to a very senior leader, I would obviously want that to look better than how it does now. I'm not sure what the product roadmap looks like right now, but I'm sure they would have something on the product roadmap. If these are not already a part of the product roadmap to improve the product, then I think they would be missing the bus.

UiPath Process Mining is not as user-friendly as Celonis. While Celonis offers a drag-and-drop functionality for building dashboards, UiPath requires some coding expertise. Additionally, the navigation in UiPath Process Mining can be confusing for beginners. This makes the learning curve steeper and the time it takes to build a process mining project longer, especially for those who are just starting. However, once users get the hang of it, the tool becomes easier to use. Nevertheless, for non-technical users or those who have not worked with process mining tools before, the learning curve can be significant. While a technical background helped me pick up the tool quickly, others may find it more challenging.

What needs improvement in UiPath Process Mining is manual delay support. For example, I have a pending workflow, and I've not taken action in the last five or ten days, and because of this, the whole process gets delayed, but UiPath Process Mining doesn't provide any manual delay insights. It gives about seventy to seventy-five percent detail on end-to-end process delays, the scope of automation, or the gap, but not about the manual delay, so what would make the solution better is if it communicates manual delay information.

Every business process relies on human activity, with either approval or review, so UiPath Process Mining should capture that area and provide a workflow or diagram that shows more details, such as the impact of the delay, the workflow time without the human delay, etc. That could be too difficult, but it's what I'm expecting as a customer. More automation in UiPath Process Mining would be great.

Since we haven't faced any hiccups or issues using the solution until now, I won't be able to provide any constructive feedback about the solution.

Regarding an area in the solution needing improvement, I want the interface to be better and easier. The solution's UI and UX could be made better.

I believe it is adequate. I don't think there is an issue with the tool.

While it is fairly stable, this is something that could be improved.

For me, the only improvement needed is how to manage the actual project. This does not have anything to do with the solution itself, it's about managing customer's expectations.

UiPath could improve on how it works with the project management process and flow. 

I would like to see UiPath add an RPA that handles project management. I would  like to see the solution do as much as possible. There are so many turning wheels on a project that is complex.

We still need to explore how to integrate the solution into various domains. It looks promising across the board. 

The initial setup can be complex. 

It's not very easy to turn raw data into actionable information. You have to prepare your data before uploading. In the pn-premises version, it would often take four hours to prepare a configuration for a single process. Today, you can do that faster. However, you have to create files with various templates - and you have to use specific formats. It all has to be extremely specific, right down to specific log names and column orders. It's a constraint.

The end-to-end visibility could be made more clear. I can't see a specific view frequency or activity frequency.

Performance checking could be improved. I'm not sure if this is due to the fact that they are gradually migrating old features from on-premises to the new cloud version. 

UiPath Process Mining needs system development and capabilities for designers. UiPath Process Mining is not easy for everyday users. UiPath Process Mining requires system development skills which means you still need special development training for all the developers. This will make it easier to use out of the box.

The ability to transform raw data into actionable information can be challenging due to the variability in our customers' processes. While online, in-system data is readily accessible, offline or manually processed data presents significant hurdles.

The pricing can be improved.

UiPath Process Mining could improve by limiting the number of updates.

One key aspect is how UiPath in itself can provide certain views of where an exception has occurred. If UiPath can identify where it has gone wrong and try to identify the bottleneck value, which are more machine learning in AI, it would be great.

If they can strengthen that core area as well as in the business processes side, not as a separate module but within the current RPM module itself, it will be really helpful. If they can bring AI and ML algorithms within it, it will be great.

Document Understanding can read up to a certain amount of accuracy. While I understand that they have already started working to improve that accuracy right now, I think the accuracy is still only around 85-90%. I would call that our main problem with UiPath currently.

Then, in terms of process mining as a procedure, it only helps in terms of whatever processes have already been digitized. In this sense, what we find lacking is that at times when the file is a photograph, let's say a JPEG file, then because of its file type, UiPath's Document Understanding does not read it.

So we'll have a situation where people scan their documents in (or take a photo of the document) and send them across in an image file and not something like PDF, and in that case Document Understanding isn't very helpful.

Process planning to utilize for streamlining processes is difficult.

The solution has its own language to process the data, so it can be difficult for people to start using it quickly. It can be difficult to handle the first project until you understand the product a bit better. 

Data manipulation could be improved. It can be very hard to manipulate or work with data sometimes, especially at first, when you are not familiar with the solution. It can be hard to learn the solution. 

UiPath could provide more AI and machine learning tools to analyze processes. 

They could also improve the UI experience. For example, how graphs and charts look and how they can be placed in the dashboard. 

The connection to other systems could be better. I don't think it's an issue right now, but I think it could be better. I would also like the connection between ProcessGold and the RPA's interface to be more transparent.

One of the main pain points that I identified with the tool is the simulation. What I am looking for is a tool that is able to help us simulate scenarios. In practice, we pose scenarios to determine what happens if we change the value of a specific step in the process. A good simulation module would help you answer the question as to how well making that change will go. Process Mining is more of a pure process mining tool, but I think adding simulation capabilities will increase the value of the solution exponentially. Adding simulation capabilities would allow us to do some additional forecasting as well. I think my suggestion — considering the product category — is doing more to support integration with that type of simulation capability even if it is not directly in the product.  

A second pain point has to do with a deeper exploration of what you have mined. If there were an option to export to UiPath Orchestrator or the UiPath Studio functionality you can then further explore and utilize whatever you have mined. The export would make the process into some usable components that could be mined or used in other ways. I think that will be a great help and addition to the product. It would mostly be useful for development purposes.  

The product could diversify into creating apps. For example, if you look at Power Automate, it's not just RPA. They also have applications. You can do low-code applications. They could add something that can help develop very small mobile and web applications, which is where the industry is going. So maybe UiPath can also follow suit.

In the current COVID-plagued environments, we built many bots, which was a plus, in place of requiring people to be physically present in the organization.

We use UiPath Mining Review in multiple forums. At the day's outset, we upload economic information and make much use of the financial reports at day's end. This used to be done manually, but the solution now allows us to do so automatically. As such, we focus our attention on many areas. 

This said, the remote desktop we utilize suffers from inconsistent performance, sometimes functioning well and other times not. 

There should also be adapters for enabling more control of the remote desktop. The solution should include some adapters for numeric use, as well as a remote tester. 

There should be better handling of errors or exceptions.

The dashboard could be improved by allowing customization and they could introduce a Mac version of the solution.

In a future release, they could provide more API integration, for example, with Oracle API.

I don't think Process Mining has any areas for improvement. I think once you start using it, it's absolutely vital for a project. It's truly one of the most helpful things for a project.

It should be an original part of the package and not just an add-on. Obviously, that makes it more expensive. But I think that once people see the value and they use it, it's so important for everybody. The BAs are really impressed and the developers really get a kick out of it — they love it. It makes it easier to sell. I believe it's even more valuable than Automation Hub.

I used to be a trainer and an instructional designer. I think the training has to be more specific — it's a little scattered.

An area for improvement in UiPath Process Mining is that its pricing is too expensive, particularly when compared to Power Automate.

I think they need to look at pricing. Since acquiring Process Gold it was stable on process mining. But now in the market, there are other process mining applications that are far cheaper. I think subsequent versions should focus on reducing the price.

The configurations on the ERP platforms could be improved. In the next release, I would like UiPath Process Mining to include better integration with some of their other offerings, more ERP platforms, and more legacy systems.

It lacks certain features. For example, if we have data from one activity that is connected to another activity or number and there is a one-to-one connection, when we want to dig into that, it fails. I had to create another workspace and make heavy changes, which was time-consuming. 

It should be more intuitive and easy to work with. There should be more ways of creating filters, and it should be easy to make our own customized dashboards. Currently, if I want to create a dashboard similar to the one we have in AppOne but with different attributes, it is very difficult. There are activities, but I don't want activities. I want other attributes. Its user interface is not very intuitive. You need to play around with it. I have been playing around with the UiPath Process Mining tool for almost two years, and now I feel confident in handling it. It is a difficult tool.

In addition, if there were dropdown options for picking an activity and dropping it somewhere, it would make the work easy.

Some things could be improved in ProcessGold even though it is already very useful. New integrations sometimes depend on new technologies or new hardware. The compatibility with the various integrations with new technologies should be enhanced so that these can be adopted as they become available.  

It is an expensive solution. It would be good if the company could consider some type of change to the licensing model that could help clients better regulate the costs in different ways.  

Another improvement would be on the development side for test management. It would be nice to be able to create new custom features for testing within that module.  

Documentation could be improved and the cost of the product is high, even if you're a middle scale industry. You might reject the solution on the grounds of cost. Power Automate Desktop is much cheaper than UiPath and Automation Anywhere is also cheaper. 

In terms of what can be improved, what I've seen so far in most of the implementations is that the client's expectations towards the RPA solution and their road map for the solution are different. The road map should be very clear in terms of what should be done and why it has to be done.

As for features that should be included in the next release, I would say that the UiPath should follow a flow X where a person who has a business process understanding and has an understanding of UiPath would be able to automate their own processes.

I would like for user onboarding/onboarding of enterprise clients to be simplified in the next release.

We have too many processes. We use non-script data. We need to see how to capture this information and do a process mining to improve the data - to clean that data and store that data but in a clean way. We still need to figure out if UiPath can do this effectively.

We may need additional features for processing data.

The interface could be more user-friendly so that it is easier for business people to use. As it is now, using this product requires a specific understanding and training. If the interface were simpler then it would be easier to use.

We found that the flow chart is not very clear for some of the complex processes. If I'm coming from A and going to either B or C based on some logic, the flow chart should depict that clearly. But in some cases, based on the logic, the flow goes from B to C, defeating the purpose of segregating B and C in the first place. So in these complex scenarios, the flow chart does not provide a clear picture. It may be due to a tool or a process issue, but we have difficulty understanding it. Therefore, this is an area of improvement for me.

One of the challenges with UiPath is the crunching of data. The solution is not so smart so we have to use a data analytics team to do the crunching for us and then import the data into the tool. UiPath should have better connectors to the systems once they integrate with SAP. They have standard SAP Salesforce integration, but when it comes to third parties, integrations could be better. There are other products that have better connectors than UiPath.

Getting a salesperson to actually call me back when I have a pricing question is extremely difficult. They need to work harder at getting back to their clients when they have queries.

I don't have any special requests in terms of new features I would like to see in upcoming releases. Right now, it does everything we need it to do.

The training materials sometimes don't keep up with the system changes.

It would be nice if UiPath's Document Understanding feature supported more languages. I don't think it supports Chinese or Japanese. We have been working in the Asia-Pacific region for a long time, and this prevents us from recommending the solution to some customers.

There is room for improvement in the design studios. Design studios and exception handling are the two that we usually migrate to Blue Prism because Blue Prism's exception handling and design studios are much more, user-friendly.

It would be helpful to make the spy modules more accurate.

The spy modules will integrate your Citrix-based programs. Many people are moving to Citrix-based systems.

In an upcoming release, they could improve on the customization.

The software at first is hard to use and you're not sure where to start. I would like a few tutorials that explain step by step how to start, including a few tips for new users.

I would like it to be more user-friendly and more visually appealing so that my eyes can keep watching for hours without getting so tired.

I would like to know what the exceptions are when they occur, including more details about them.

I would like to see enhancements to the recording feature.

Compared to other technologies, the initial setup can sometimes be a little more complicated. It can be difficult to choose the best way to do something so maybe they could improve the roadmap. 

I think the user code feature could be improved. 

They should work to make it possible to handle legal processing tasks easier and more accurately.

There are a few complexities involved in setting up the solution.

We do not have much confidence in the intelligence or cognitive capabilities and they should be improved. Automation Anywhere is really good in that area, so I'm looking forward to more advanced features in the future.

I would like to see better integration with the Azure and AWS machine learning platforms.

It's quite an expensive solution. It's not competitive right now. 

It's too early to say as we've only been working with this solution for six months. However, the main issue I have is with its programming. I think they need to work more on producing low code to make it more end-user-friendly. I think there's a big dependency on developers and technical teams in the background.

The main issue for us is how data is being collected. I think a lot of Process Mining's tool has a very limited way of collecting process data. Industry-wide there's a gap and data collection needs improvement, it's a constant challenge. Any demo will show an ideal case but when you use it in the actual environment, collecting data is a real problem. Cleaning the data is also a problem, and then being able to analyze it and make any meaningful sense out of the whole thing is the issue. 

The main thing that can be improved is the price, nobody wants to pay that much money for a scheduler.

Another feature they can add is, instead of doing only UiPath bots, it should handle other bots as well. Then I can deploy automation anywhere on any other software. It should not be only UiPath. That's the future of it.

We feel there is some problem with the user interface, it is sometimes difficult to navigate to the area you want. Additionally, the bot hangs at times.

In the next release, it would be beneficial to have a conversational bot that can give voice alerts for problems.

Licensing could be cheaper, it's quite expensive. 

The user interface could be more user intuitive.

Also, they need some better data management features. For example, as far as what we are collecting, being able to decipher and comprehend what the process or flow is, that we are looking at. 

They could increase the feature set.

I would like to see the ability to do fixes, such as to address and mitigate some of the gaps or issues in the process flow.

They could develop models, or use cases that can be leveraged for mitigating and remediating process gaps.

Process mining doesn't need an improvement as such. However, to get the data from the business process, the backend data, and the log data, from the business and feeding that to process mining is a challenge. 

Not everyone in the business can readily provide the kind of information that is to be put into process mining, which is an area that needs to be worked on.

At the moment, process mining is a new product, so the pitfalls or the cons of this tool are yet to be realized. It's too early to say.

What I would expect is a little bit of an AI component included in process mining. That would help the developers or the business analyst get more insights into the other possible options. 

The element of features that can be included. At the moment it is data-driven. There should be some improvement in terms of its intelligence to artificial means. That is what I'm looking for from them.

There is always a challenge with the infrastructure set up for the client. Setting up the server is something that might be difficult. It needs to be simplified to make it easier to install the tools and start working on the automation.

This is not a perfect tool, as there are things that should be improved.

Process Mining still has to mature somewhat.

There is a bit of a concern about being able to use this on a mobile device.

If somebody opens an Excel book and they didn't maximize it, you have to open it in a maximized mode, the occupancy then will be identified properly by the pod which is opening. If you're using DL, it works fine, but sometimes our requirement is to put in a CC.

Improvement depends on whether I can get the input information to explain why the exceptions are happening. Is it because of different policies or incorrect procedures being applied? If that information would also be integrated as part of it, it would make my life easier.

How the input is received from the medium and assimilated is important. The kind of application that we are using to process the particular process also plays a role.

I would like to know what the exceptions are when they occur. Is it because of the process selected needs or is it because of policy, legal issues, or incorrect procedural obligation? That would really help me.

In general, if they can provide the exceptions, as well as why those exceptions are happening from the input side and what applications had been used to process the information, it would be much easier for me.

Its license structure is complex. 

In terms of features, there isn't anything missing at this time.

I have found UiPath Process Mining to be a very complex tool to use. They could improve the OCR recognition feature.

I would like to see more enhancements to the recording feature. Also, more logic-based conditions and flows should be shown.

Its price can be improved. They can also have a better plan for training the end users to be able to use this solution.