WorkFusion Reviews

Within my company, I am responsible for a line of business where we roll out automations for various internal stakeholders, and WorkFusion is the platform of choice within my organization.

Approximately 90% of our use cases are RPA-governed use cases. We have one use case now that is a machine learning (ML) based use case. While we are not adverse to using ML within WorkFusion, we are still in the process of identifying or exploring the right use cases within the organization that are the best fit for ML.

A helpful feature is the analytics capability within WorkFusion. The analytics dashboards, i.e., their Tableau dashboards, are really helpful for building a business activity monitoring layer. These are very visible to senior management and stakeholders, who can see and know what is happening within processes. They can see what benefits are getting delivered as processes, either ;daily, weekly, monthly, or quarterly. It is easy for stakeholders or senior management to understand the value for the money that they are spending.

The solution has ELK Stack with Kibana, which is useful and helpful for monitoring all of WorkFusion's infrastructures. This has been really helpful, even though we have just started using it very recently. Previously, if there was an issue, we would log into each of the servers separately and try to fix it. We were introduced to ELK stack two weeks ago, though it has always existed in the product. It has been really amazing. If there are failures, we immediately get notifications. We have been able to integrate it with our enterprise alerting system that creates tickets automatically, which is amazing. We don't have to spend and dedicate hours just to look into the infrastructure to see if there has been a failure or not. We can be reactive instead. 

WorkFusion definitely has a very strong capability to process any document using OCR and machine learning frameworks, particularly from an extraction perspective. We have just used this, deploying one use case using these capabilities. We have been happy with the results. We are getting almost an 80% automation rate. Based on this success, we have started exploring other opportunities within my organization to see if there are other use cases that best fits this AI paradigm.

It is difficult for non-technical users to use WorkFusion. If we really want to build automation, we definitely need Java developers. Their RPA is not for non-technical people. If I look at competing products, like Automation Anywhere, UiPath, Blue Prism, or even Power Automate, a non-technical person can build basic automations and deploy them. Since non-technical people can easily build basic automations, they can be deployed in no time. With WorkFusion, there isn't a similar feature that works well. One needs to write code from scratch to build automation. That is definitely a deterrent if we really want to do quick, dirty automation. I would not rate WorkFusion very high on their no-code, low-code paradigm. In this area, I would definitely expect them to improve.

From an AI engine perspective, they are currently focused on document processing. They should provide the capabilities to integrate ML models. which are just outside document processing. In the world of machine learning, there are several such models available. They should give that capability or provide those capabilities within the platform. That would go a long way.

I have been using WorkFusion for more than six years in various capacities. 

Since we are working primarily on RPA, I find their RPA nodes to be not very stable at times. The nodes may crash at times, and it doesn't send any notification that the nodes have crashed. Maybe the configuration at our end is not right, but this is what we see. From that perspective, we need to continuously monitor those nodes. If we had a better way of monitoring all those available nodes, it would be much better. That is the most fragile part of WorkFusion's infrastructure. Every other component within WorkFusion is very stable.

RPA, as a whole, is an amazing product within WorkFusion, when it comes to deploying at scale.

WorkFusion is the platform of choice for scalability. This is one reason that we are using WorkFusion compared to other competitive products, where the ease of building RPA is much better. If I look at the competitive products from a scalability standpoint, WorkFusion is the platform of choice for us. It is amazing.

We have already 50-plus processes deployed in production. We have around 50 to 70 more processes in our pipeline, and the pipeline is growing.

We have a 20-member team at the company working on WorkFusion.

From a customer support perspective, we expect timeliness of response. If we are raising a ticket, we expect that whoever picked up the ticket will respond in no time. At least, the first response should come in no time. After the initial response, if they require certain logs, then they should take them from us and come back to us within a stipulated time frame. It should have the right recommendation and a solution that should solve the problem.

Their response time is sharp. If we raise a ticket, which is a production-critical ticket, their initial response comes in less than 20 minutes. I haven't seen any tickets open for more than 48 hours before they are getting closed, however complex it may be. Their support team is doing a wonderful job. I would rate them as eight out of 10.

The initial setup wasn't complex. They definitely expected us to provision several infrastructure components, like servers. However, once the servers were available from the organization, all the ports were open, and the DNS names were set up, then the installation of WorkFusion was smooth. It takes anywhere between a day to two days at max.

It took us almost a month to implement their machine learning models for OCR and Intelligent Document Processing. This wasn't WorkFusion's problem. It was more on the data collection side. Their model requires us to bring in a lot of documents to train the model. It took time from our side to provision those documents with the needed quality. The model doesn't work if you give it 10 documents. The model won't suddenly launch. It doesn't work that way. Once we had given it the right set of data, then the initial out-of-the-box model was done in a month's time.

When we were building our first ML model, we obviously weren't able to achieve the success rate that we were aiming to achieve. We engaged WorkFusion who spent almost a month with us unconditionally, helping us with our complex document structures. The results have been amazing. 

From an engagement standpoint, we are happy with WorkFusion. Their customer success team, delivery team, and sales team are amazing to work with.

We have not yet seen ROI. We have been using this product for two years. We realized that we need to deliver more automation on this platform to get the benefit. That is a function of the use cases that we are picking up. Our success criteria for automation is a reduction in FTEs, which is because we are measuring a break-even from a FTE reduction rate. This is also a function of how we are internally finding out use cases within our organization.

The pricing is high. They should have a better pricing structure. It was not easy for me to try out WorkFusion initially. For example, if I am working on SAP, the licensing may be very expensive. Though, it is fine to use it at scale and pay for the license. We have an enterprise license now, so we are not worried anymore. We are fine with it. 

As a first-time customer who wants to engage with WorkFusion, not working for my company, then I would want to do a PoC. The cost is prohibitive because they will ask me to get infrastructure with six or seven servers. Then, you must buy a WorkFusion license to do a PoC, which is expensive. I would then have to set up a team to build the PoC and see the benefits. That is very prohibitive.

We explored other solutions, like UiPath and Automation Anywhere. We even tried Power Automate. We evaluated them almost two and a half years ago, then we settled with WorkFusion as our platform of choice due to its scaling capabilities.

When it comes to building automation using RPA, it is not very intuitive or user-friendly. We definitely require developers to build automations. Compared to using alternative products to build automations, such as Automation Anywhere or Power Automate, WorkFusion is not easier for building quick and dirty automations. However, if you want to build automation that can be deployed at scale for several transactions within a day, WorkFusion's RPA is amazing. This is one product feature that we found rather interesting.

If you are looking for building small-size, quick and dirty automation and something that you can build from scratch within two to three days to run on each of your user's desktop, then WorkFusion is not the platform of choice. However, if you are looking for building automation which can run at scale, e.g., if you are doing suspect transaction processing in a bank and processing approximately 50,000 transactions in a day, then you want to really ramp that up at scale. In that case, WorkFusion is the platform of choice.

There is room for improvement. I would rate them as six out of 10.

WorkFusion is an RPA solution that we use to automate things. We have multiple use cases. We use it to help prevent card fraud, as well as for fraud investigation, credit, and new-account opening. It is used across the bank for multiple purposes.

The purpose is to automate trivial and mundane tasks and make them less dependent on humans, as well as to make them much faster and more efficient than when done by humans.

The whole purpose of RPA is to minimize and automate the mundane tasks that go on in any organization. In our case let's take the example of card fraud. When someone here in South Africa suddenly gets an SMS on their phone saying that 1,000 South African rands were debited on their credit card, and they realize they didn't do the transaction, they'll report fraud to the bank. Before automation, when someone reported fraud, the case had to be logged and an investigation would take place to look into whether the customer had actually been a victim of fraud. That process included pulling SMS and ATM logs and it took seven to 10 days for a case to be finalized.

With automation, we have some pretty new solutions wherein, as soon as a card fraud is reported to the bank using any of the usual channels, whether it's through the app, an SMS, visiting a branch, or contacting a call center, a case is logged and flows into the robotic system. Then, based on the customer's card number, his details are fetched from the bank systems, including his SMS and ATM logs. When an investigator sees that the customer was a fraud victim, he puts it into the refund queue and the bot reverses the amount. If it is the first time a customer is a victim of fraud, he gets a refund within 10 to 15 minutes.

We already have a lot of compliance and governance systems in the bank, but they were all manual. If someone who was doing compliance and reporting was sick for two or three days, that job would be pending or maybe someone would need to take over. With automation, the level of reporting has increased because we are doing all the user reporting more frequently. WorkFusion has increased the level of compliance by not delaying it.

The Control Tower enables us to create and use workflows. It helps us run processes. There are a lot of small features, but the Control Tower is a very good feature, wherein we can create bots and schedule them as well as create different instances of them to run in different regions.

The "human task" feature is also important. It allows us to handle scenarios that might be causing a process to fail. We can catch a process that has not finished and take action on it.

When we develop bots, WorkFusion is heavily integrated and built on top of Selenium. We get to use the Selenium capabilities, which is also nice.

I've not worked with the OCR feature, but from what I hear, its accuracy is not that high. It depends on the type of documents you're processing, because if they are handwritten documents, it's very difficult to do optical character recognition. Maybe it is a matter of using the AI/ML features of the WorkFusion platform, which we haven't explored. But document processing with respect to OCR can be made better.

Also, WorkFusion comes packed with certain libraries and .jars. At times, you understand that you're using a system in which a library or .jar is outdated and you need to upgrade it. At that moment there is a problem because the WorkFusion guys say that you cannot tinker with their basic product, despite the fact that we have encountered an issue with an existing library or .jar being either outdated or having some serious issues. And because of that, we are failing. They need to better manage quickly making changes in WorkFusion to accommodate a situation where other technologies are not supporting the platform.

For example, WorkFusion uses a protocol for a file transfer protocol server. Now, some other systems we are using WorkFusion with have migrated to a higher protocol. WorkFusion doesn't support that higher protocol and the other technology doesn't support the lower protocol. We are in a Catch-22 situation. But if we were to go to WorkFusion, they would say we won't allow you to change our product. They can do better in managing changes customers need to make because of certain situations.

I have been using WorkFusion for about four and a half years.

I actually never worked on RPA before I started working here. I had knowledge of Java so I was hired by the organization. There was a lack of WorkFusion people available in the market five or six years back. Even now, it's not one of the most popular or prevalent technologies for RPA in the market. So there are fewer people to work on it.

The solution is pretty stable. At times we have problems wherein our platform team needs to restart their services or server. But I would assume that it is because a lot of teams deploy to that server. So if they're not managing their processes in a smarter way, at times it gives them problems. But otherwise, the platform is stable.

We haven't scaled it on a dynamic basis, in real time, as you might do with Amazon Web Services or cloud services. But if we have a bot working on four machines, and we need to handle more loads, the very simple way of doing it is to add more machines or VDIs. Another option is to make them run for more hours so that they can complete the load. We are able to manage our loads with our design and the way how we implemented the solution. We haven't had challenges.

Currently, we have version 9 which is on-prem. From version 10 it is on the cloud. It's a big move from on-prem to cloud. So a lot of people have to learn the basics of the cloud. Then the whole program can move to the cloud. There are discussions going on and the next move is to the cloud.

There are a lot of people using the solution within our company. Some are technical people who are developing and deploying solutions. Others are on the platform team and they are managing the end product. Some people are analysts who have different levels of access. And some are on the business teams in the bank, who actually run the bots. There are six or seven teams involved with a total of about 100 people using it.

Sometimes, when we report an issue such as a possible bug in the WorkFusion platform, and we need answers or a solution, we think WorkFusion can do better. Most of the time, there is a lot of back-and-forth communications and then they will say, "We will log a ticket and it will be in our queue and we will investigate it." Many times it's a burning issue for us. The support from the WorkFusion team is slower than what is expected by their customers.

There could be many reasons for it. They develop a lot of features and they have to integrate improvements into their solution, and they have to roll out patches. But, when all is said and done, it would be better if they were faster and more nimble to accommodate changes.

Positive

In our organization there is a platform team that manages all the technical aspects of installing, managing, and restarting services.

Maintenance of the product is typically handled by that team as well, which has five or six staff members. There are two senior guys and three or four very junior guys who just joined out of college. But obviously, their job is not only to maintain WorkFusion, that's one of their jobs. We don't need a team of five to maintain WorkFusion, but we have a team that does the job.

It competes with other solutions in the RPA domain, such as UiPath, Blue Prism, and Automation Anywhere.

We haven't used the AI and ML features in production that much. The OCR features are reasonably good, not very good, but reasonably good, and they get some use.

If you're an analyst or an end-user, it's not very easy, but it's worth developing some abilities based on basic certification or training. It's not that complex because there are so many facilities available, but if non-technical users are not aware, they may only be able to use it to a lesser extent. It is not really easy for non-technical users, that's what I hear, but it is worth it. It gives you results if you learn it. And for a technical person, it is Java-oriented technology, so you need to have good skills to develop bots.

I started with invoice processing. That uses a lot of different parts of the WorkFusion solution. It uses their built-in OCR package, built-in S3 instance, and built-in databases. Invoice processing is the big use case. 

We use their AutoML tagging machine learning use cases. Then, at the end, we use their true robotic process automation, using the SeleniumLibrary, to enter things into our ERP system.

I have used it for a couple of other projects. I have done this mostly because WorkFusion makes it easy to code out a straightforward business process, put it on a schedule, and hook it up to a database without needing to requisition all new things and servers to launch it. I can host it in one central spot.

Our particular instance is on-prem. It has five or six different Linux servers that all talk to each other as well as a handful of Windows machines that end up talking to the central Linux boxes. In my time here, we have had two different major versions of WorkFusion: 8.5 and 9.2. We are currently on version 9.2.

The workflow automation has definitely been helpful. It gives us an interface to launch code on one site, then have it run in parallel one to ten times on ten different robots. They can do the same process ten times in parallel on a website. 

We have entire workflows predicated off of the document processing engine. Our whole invoice ingestion process runs 50,000 documents a year, give or take, completely predicated on WorkFusion's document workflows.

We receive invoices from 50 to 100 different vendors. They all have different formats. The document processing engine (coupled with their AutoML services, data extraction, and human tagging) is able to translate these 50 different formats into a single output that we can then pass downstream into whatever process needs to receive it. So, this has been a helpful integration.

We have automated a lot of finance workflows using WorkFusion. This definitely helps both our accounts payable and accounts receivable teams.

I really like the machine learning and tagging. They give a decent UI to allow even non-technical users to gather or tag the data, then that goes into the back-end so we can use it to train the models. This is something I have not seen in another solution or environment, and having that built-in, straight into the package, has definitely been a big help.

Out-of-the-box, the machine learning in the document processing engine needed some tweaking. A couple years ago, another engineer at the company and I met with one of the WorkFusion ML engineers and built out a custom model on their AutoML platform, then deployed it in January or February of 2019. It has been running pretty well since then. So, one of my huge selling points on WorkFusion is the AutoML.

We only have our engineering team writing code and developing business processes since it is very technical. In terms of data extraction and training the ML models, we deployed that to our accounts payable department, which has fewer technical users than our engineering team. After a handful of training sessions, they were able to get in there and use it with decent accuracy. It could be better, but it is definitely better than nothing.

One of the big upgrades that I know is happening in version 10 is an improvement to their tagging interface. That is definitely something that I am excited about to sort of reinvigorate our more non-technical users in information extraction from our documents.

There could be better version control through their web platform. That is something that could be improved on. Integration from the IDE, e.g., from WorkFusion Studio directly into Control Tower, would be nice to have too.

I have been using WorkFusion since my first day at the organization. The organization obtained WorkFusion sometime in Spring of 2018. I joined the organization in Summer of 2018. Therefore, I have been using it for about three and a half years.

The vast majority of the time, there are no issues. Most of the issues that we have run into have been more around certificate upgrades and stuff like that. That is probably not WorkFusion-centric and could be mitigated more on our side by using a cloud solution instead of on-prem.

The ML server specifically likes to lose connection with our network drive once a month or so. However, I don't know if that is actually a WorkFusion issue or if that is our network team doing a server reboot which just doesn't reconnect automatically. Nine times out of 10, WorkFusion has been up and running exactly when I need it.

In terms of web automation and desktop automation, it is mostly a server provisioning thing. That is mostly on our side. Once we have the servers provisioned and imaged, we can deploy one to 10 processes that follow the same code with relative ease. I have never really tried to scale beyond 10 or so.

The technical support is super responsive. Whenever we open a ticket in their Jira, they get back to us in short order. They are always knowledgeable about what they need to do when helping us fix our issues.

I would advise using WorkFusion's support whenever questions come up. They know what they are doing and read the documentation thoroughly.

I would rate the technical support as 10 out of 10.

Positive

Three years ago, you could upgrade from version 8 to version 9 to get X, Y, and Z improvements. Now, we are at a point where we can upgrade from version 9 to version 10 to get A, B, and C additional improvements, and it is good that WorkFusion is iterating and developing their solutions.

I was fairly heavily involved in the version 8 to version 9 transition a couple of years ago, and it was a big undertaking. That is something that they could do a little better. I would recommend having them assist in transitioning their platforms through the major version upgrades. A lot of things had to be rewritten in the new version as opposed to having some sort of import tool to move them over automatically.

Going to version 9.2 was not a direct upgrade. We actually provisioned new servers and did a brand new install onto them. It definitely leaned more towards the complex side than the straightforward side.

I remember our manager at the time spent a lot of late nights back and forth in the WorkFusion documentation. They have six or eight different servers that all have to talk to each other under the hood. They have to share secrets between each other to manage it. They have to manage certificates against each other. It is just the tech stack complexity that influenced it a lot.

The OCR implementation was straight out-of-the-box. 

There was one instance where I connected with WorkFusion over the development of our custom model.

Leveraging the more AutoML and custom models has been a lot of tweaking over the years. The initial model probably took about a month, once we went live with version 9.2, and that was working heavily with the WorkFusion AutoML engineer who is associated with its deployment.

The 9.2 implementation was mostly done by our technical manager. There were also about three engineers tangentially involved, mostly as a part of learning more about the solution and the inner workings of it.

Their robots have increased the rate of efficiency by two to five times in our organization. There is still a human element to code out the package running the robots. You can't just click something and have it run. However, once that has been done, you can parallelize it 10 or 100 times in a fairly straightforward manner. 

When we are automating things, we are just completely eliminating the chance of human error happening. The computer does it, and the computer will do it the same way every time. We don't have the element of humans in there. It has improved our accuracy and compliance since we are eliminating the human error element.

It has provided time savings and let employees be allocated to other job tasks.

Their OCR package has been alright. I have played with the Microsoft Computer Vision packages, and this tends to do better on certain things, like handwriting analysis. In terms of ease of deployment and integration into the rest of the platform, it definitely provides advantages with ABBYY, which is what we use in WorkFusion under the hood.

We have explored some Microsoft solutions, but they aren't exactly complete parallels to WorkFusion. They just allow similar things. Where they often fall short is ease of deployment and integrating separate parts. For example, in Azure, I can set up a database and a process that will read from the database trying to link that into a machine learning algorithm. Linking that again to a document tagging type of setup would be tricky to do. WorkFusion allows for relative ease when linking all those separate portions.

I have been fairly happy with it. It gives me straightforward deployments of stuff. It lets me build out schedules and set up database tables natively from their web interface.

I led our internal 'tagging' training sessions, not WorkFusion. It was mostly just walking through what the environment is and what they have to do in the solution to process documents.

If someone was looking into WorkFusion, but already has an RPA tool in place, I would probably want to learn more about what they are not getting from their current RPA tool. Then, I would figure out if that is something that WorkFusion would be able to provide or if it is something that would be a shortcoming in WorkFusion too. Generally, I would lean towards the side of momentum, where if they already have a good thing going with one tool, unless there is a real need to transfer over, then I would recommend staying with that tool.

I would rate the product as seven out of 10. It is definitely not a nine or 10 because there are certain shortcomings in it. However, I don't pull out my hair every day trying to make general things work the way that they should work.

We use WorkFusion for SPA, Smart Process Automation. SPA combines robotic process automation and machine learning. The heart of it is machine learning. The RPA processes are a few steps that a robot does instead of outsourcing to a human in India. It makes the process more efficient to execute any exceptions on the WorkFusion cloud.

It's a SaaS solution, so everything goes over the cloud. Only the results are delivered locally on-premises. We're using it for machine learning purposes to extract data. However, the process is not entirely straightforward and sometimes we need manual intervention by our third-party ops provider in India and Poland.

We have between 20 to 40 bots that work in our production environment. They do the work in the background, so we don't see what they do, but they handle the workload. We also have approximately 20 employees in India that for whenever a bot has a question or has an issue. Then a human being needs to look after that. This gets done by our colleagues in India based on training. We define the gold data to be specified in each document we want to have extracted. We also have very detailed instructions on what has to be done if something happens that is not as expected.

We work directly with WorkFusion. We also work with a back-office services provider and they do the manual exception handling. Whenever something isn't automatically extracted, it appears in the cloud. An employee of our service provider in India or Poland handles the manual extraction based on the detailed instructions. We get the results and import them into our target systems.

WorkFusion has doubled our efficiency. We replace manual processes with automated ones. We need half as many people to process the same amount of work, or we can process double the amount of work with the same people. It's a mix of both because the volume has increased a lot. 

We would have some issues meeting our obligations to our clients without WorkFusion. We expected the volumes to increase, but they coincidentally spiked as we introduced WorkFusion. That was an immense help because we would've had trouble serving our clients if we didn't have this capacity gain.

The document processing engine is essential because that's where we gain speed. We're processing a massive volume of documents with the document processing engine and automatically extracting around 80 percent of the information, enabling us to honor our clients' SLAs. We can distribute the information 24 hours after having received it. This is an SLA that we have with the clients, and that document processing engine helps us keep those timelines because it's a lot faster than if someone had to look it over manually.

Adopting WorkFusion has improved our accuracy, but not significantly. We have fewer mistakes overall compared to parallel processes that don't go over WorkFusion. It's around 5 percent better. Compliance isn't an issue because we process publicly available information. It depends on how compliance is interpreted. We would also have more security tasks to go over internally before working with an external provider.

When we implemented WorkFusion, it was quite popular within our company. I got to do a lot of demos for different companies. It was good because it showed people what was possible and inspired ideas about new use cases. I'm currently also working with all kinds of different use cases. In terms of creativity and generating ideas, WorkFusion was good. 

We're also able to automate internal processes because WorkFusion also can deliver different outputs that are more flexible for us to work with internally. So we're also able to have more efficient internal processes thanks to WorkFusion.

The document processing engine is good. We're extracting around 70 data points from a given document, and we have an automated extraction rate of 80 percent for those fields. For some fields, the rate is 100 percent because they're clear. They're always in the same spot and are very defined. Other information is only found in a few documents, and the automated extraction rate for those cases is a lot lower. Overall, we're quite happy with the 80 percent.

The interface is user-friendly and accessible for people without a technical background. It's straightforward to design and set up the business process ourselves. I think the UI was designed for business users rather than IT or technical users. I was able to set up a business process with a half-day of self-guided online training. That's pretty intuitive, and I would rate the UI 8.5 out of 10.

I would even give WorkFusion a nine for overall ease of use. The workspace is well designed. You can navigate everything using the keyboard, and it has good functionalities. WorkFusion is ready if you have specific improvements to your business processes. 

The most common issue we have is with the OCR, but that also depends on the quality of the documents your scanning. Still, I would say the OCR is comparable to other solutions. I wouldn't say it's better or worse than other OCR engines.

WorkFusion entered the market early, and we see some newer providers that run on different technology in the background, so I'm not sure if WorkFusion's platform is always the most advanced. At the same time, I'm confident that WorkFusion is also working on releasing a new generation platform when they see fit. WorkFusion is constantly developing. 

However, we've had use cases where we decided not to employ WorkFusion because it lacked capabilities that were important for the business processes in our company. We discussed this with WorkFusion, and they were open about these functions not being one of their primary capabilities. In these cases, we chose another provider. 

WorkFusion is willing to listen to new requirements and look for solutions. They're also forthright enough to say when a specific capability may not be the best fit, and they're able to admit that. This is also something I appreciate because other providers claim to do everything, but when we do the POC, we find out the solution falls short of our expectations.

It's very possible that WorkFusion can address the internal RPA tool issues. But in terms of RPA, it depends. If you identify the need for an external solution, then you can do an RFI with WorkFusion. WorkFusion will give you a truthful reply to your requirements about whether they can help you out or not. We have RPA internally, and we do most of it internally. When we don't need WorkFusion, we prefer to do things in-house. We can solve most of the processes we have with our internal capabilities.

We started working with WorkFusion about five years ago. I switched departments around two years ago and since then, I have not been responsible for the original process that we had in production for WorkFusion. Fairly recently, I did a proof of concept for another project with WorkFusion.

WorkFusion's stability is fairly good. We have to open support tickets every now and then because something breaks down or isn't working. When we started working with WorkFusion, they advised us that they hadn't worked with this much volume in the past. We were processing more documents than their system had ever handled. We pushed their boundaries a little.

We weren't expecting the volume, either. When we went live with WorkFusion, the volume started spiking simultaneously. We had even more volume than we originally anticipated, and that caused some stability issues. 

Fortunately, that never caused a long interruption in production. We have an agreement with WorkFusion, so they would always address any ticket in a timely manner. We're not seeing 100% stability in uptime, but we always get support when it goes down. They have improved a lot over time.

They addressed any scalability issues we had at the time. It wasn't perfect from the beginning, but WorkFusion always showed us support and commitment to helping us succeed. I would rate them an eight of 10 for scalability. If it's not ready to scale, they will find a solution for you. They scaled it for us in the end.

We process an average of 1,000 documents a day. These documents are about 30 pages on average, and there are about 50 to 70 data points per document that can be extracted.

WorkFusion's service is what sets the company apart from competitors, which are all similar in terms of features. They're set up fairly well to respond quickly if there are errors or you need to make some adjustments. 

They also understand our use cases well. That helped us a lot because we're a financial company, and we appreciate their expertise. They understand the issues we're trying to solve. We appreciate that. I would rate their customer service nine out of 10. 

Positive

WorkFusion was the first process automation provider we implemented. 

The initial setup was complex for me because I didn't have much knowledge. I don't have a computer science degree. I had half a day of online training and did weekly calls with WorkFusion where they answered my questions and offered solutions to problems I had. 

It all goes back to customer service. It took about three to four weeks that WorkFusion.

Setting up business processes on the WorkFusion platform was painless because they provided good support. I didn't configure the machine learning models myself. We gave our business process requirements to data scientists at WorkFusion, and they implemented the background ML models. We specified how a certain data point would be recognized and interpreted in a document.

WorkFusion's automation is excellent from a cost-benefit point of view. It's not for free, but we doubled efficiency. You can look at it in terms of opportunity costs because if we didn't have WorkFusion, our costs would increase by around 30 percent.

The price for WorkFusion is quite fair. It isn't the cheapest solution in the market, but it's very close. We had a bad experience when we chose a cheaper provider for a different POC who failed to meet our expectations. That could also be our fault. Other providers will charge you more for the same amount of work.

We evaluated some solutions, but it ultimately came down to price. WorkFusion was fairly early in the market. I am not sure if their technology is as sophisticated as some of the new players on the market, but the new players on the market are also a lot more expensive.

WorkFusion has standardized processes, but they've also developed new templates and machine learning models for us. Overall, I would say they're fairly standardized. Another provider develops machine learning models on the go based on a template, but a different business model enables them to do that.

I rate WorkFusion nine out of 10. WorkFusion's customer service is the most important aspect for me. They have people in Minsk and London, so they also cover different time zones and can help our colleagues in India when they need help. One of the reasons we chose WorkFusion is because they had full-service coverage of our different time zones.

My first use case was for reconciliation purposes. In our bank, we had recon challenges with one of our tools. We built a bot that can reconcile the bank's main push-pull account. It does the reconciliations and picks out all the exceptions, things that don't match between the bank's account and the customers. For example, if there's a debit here but we don't see it corresponding there, it flags it as an exception and, at the end of the process, it triggers an email to the team that owns recons to be handled manually. But the bot does the main part of the reconciling.

We came up with bot processes that are saving the bank time. It would take days to do a reconciliation of the bank's main push-pull wallet. Because of that, sometimes when there was a mismatch, nobody picked it up on time and the bank would lose money. Now that the recons are happening on a daily basis, it's helping the bank. Now, the bank is diving deeper and wants to go more into automation.

It frees up time. The feedback we have gotten from the businesses, now that we've been able to automate their work, is that they're able to spend more time with clients that require their attention. There are some customers that would prefer you spend more time talking to them, advising them about products and services. Now, they have ample time for that. 

Somebody who previously spent about three hours out of an eight-hour day doing reconciliations was only left with four or five hours, which didn't give them ample time to spend time with customers or pick up other things that required their attention. Even if it's something that is not fully automated and requires the officer in charge to manually trigger that process for it to run, once that person has triggered it, they can go about their normal routine. It's something that some of the business team members have come to accept and they are diving into it more.

It has helped improve our accuracy because when it does the recons, it does the matching. Recently, our recons team was audited, and when the auditing was done, the process done by the reconciliation bot was the only process that was passed by the auditor. The ask now is to automate all the other processes we have yet to automate. I'm confident when it comes to accuracy because of the audit results.

The tool itself is easy. I started using it without really having any coding experience, and it was something that drew my attention toward development. I was from the business side, generally, but now I find myself in the engineering field and doing more. But it started from there, where I didn't have any knowledge. So the fact that it was just a matter of "click here, click here," allowed me to mimic exactly what I used to do for account reconciliation purposes. After just a few clicks, it was doing reconciliations for me. It is the tool that got me started on this journey because it was easy to use.

I came fully from the business. I didn't have any technical background and actually did Banking and Finance in school. I just knew how to use a laptop, but I didn't have any technical skills when I started this journey. 

For me, where it can be improved is the speed. I built two bots, one with WorkFusion and the same thing with Python. My feeling was that the speed at which Python did things, how fast it did them, was quite cool. It was faster. Sometimes, when WorkFusion is running, the way it triggers on its own, it takes time before it even launches the page. But once it launches the page, it moves. WorkFusion could improve how the tool itself triggers some of these bots on its own. I feel that it's slightly slow.

They are all fast but the other one was a bit better than it was in WorkFusion. They are doing better when compared to the old versions they used to have. When I started, the tool was way worse. The more recent one I'm using is notably better in terms of its speed.

I started using WorkFusion in 2020. I did my own personal learning around it at the start and then a certification program and became fully certified in 2021. I am now a Power User of it.

I've gone beyond just WorkFusion, and I've done a deep dive building robots in Python. I'm currently also doing that with PowerApps from Microsoft.

I feel that the slowness I mentioned is coming from it not being stable. I've had scenarios where I run a bot the first time and it gives me an error. I don't make any changes to it and run it a second time and it goes through successfully. I can't explain what causes that. Maybe it is network-related.

But in general, once we have developed the bots, they run on their own. The only time I've had a challenge where it has not been able to complete was in the event that the bot was not able to access some of the platforms we use because a platform was down. If you assign the time, "Run at 6:00 AM," by 6:00 AM you will notice that it will at least attempt to run. If there's an error, it will alert you. You can see that it couldn't log into a page because the page wasn't there, but it attempted to do it. It won't be the case that it didn't run because of something on their side.

They usually give us direct support because our bank has a direct contract with them. If we make a suggestion about what we want, they find a way to put that into the system for us. But I haven't had any reason to go to them for support. The tool hasn't given me any headaches.

Comparing WorkFusion from two years ago with WorkFusion now, there's a difference. They keep adding more. Previously, the number of pages you would get for OCR and machine learning wouldn't be that huge. You would use them up and you couldn't do anything more. And these are bot processes that need to be running on a regular basis. But now, that has been increased heavily. 

There are more additions and features being added to help make our work much easier. Before they do an update, they send emails to some of their regular users to ask, "What do you think we can bring on board in our next version that will help the tool?" They do try to release versions that can help.

Previously, in the bank, we didn't do much automation. It was a journey the bank decided to go on in 2020. The bank wanted to start automating most of its manual processes related to a type of tax. We have neighboring countries that had similar projects. The whole idea stemmed from South Africa. They were doing most of their automation in WorkFusion so we decided it was something we could also look into.

I saw WorkFusion online somewhere and somebody also mentioned it to me when we were at a session about something else. They said, "Oh, he did some automation with WorkFusion." I was a bit curious and read about it and decided to download it. Nobody guided me through it but they have online training, so that's where I went.

I went through the "How To" on their portal. They started me off by playing around on a website, capturing data, taking a few figures from online, things that had to do with rates. When I did that bit of work and I picked up the idea of how it works, I built my own quick bot. It wasn't anything huge, but it solved one of the problems I had at work. And that's what led me to further go and get certified in it.

Based on those of us who took the early lead, and our success story, the bank felt that it should invest in this channel.

You usually will use the tool itself on-prem. But to have it run fully automated, you put it on the cloud on their server, where it's able to run, connect, and then run on its own.

I followed the steps and the setup was quite okay. Even when I was about to purchase a license and switch it to their cloud site, I was able to do it on my own. It was just a matter of reinstalling and selecting the right option. Based on the videos that I was watching, it wasn't difficult at all. And that's especially true if you are a first-time user of the platform. If you watch the video, you can follow it through and then do your own setup. And you will have your own studio to do development in.

I started using the OCR after I was fully licensed. It wasn't difficult for me. You can practically show it what you want it to do on the page. You show the bot, "Go here, and whenever you see this reference, pick the next value after that." It can do it for you.

We've gone beyond just WorkFusion. We have added more tools to our development. We do Python and Microsoft Power Automate as well. We keep exploring and adding more tools that we can use internally to cover things. One thing we have noticed using some of these tools is that some things may be faster in Python than in WorkFusion. We try to compare them and see which one solves the particular problem we have and which is faster.

We attempted OCR in Python but we weren't getting the accuracy we wanted. We brought in Power Automate and noticed that there's an OCR piece in the PowerApps and we are looking at whether it can give us the accuracy we want. But for now, when we have documents and files to read, the tool we go for is WorkFusion.

If you have any technical background in Java or other development languages and you want to experience some light coding or just know a little coding, WorkFusion is good because it allows you to do a lot by clicking. It's good for somebody who wants to start this journey. There's a coding part of WorkFusion for those who want, but if you want to build your skills and start automation, it's not a bad tool to start with. In time, you're going to want more, but for a starter, WorkFusion is not a bad tool at all. You can click and it does image recognition and more. It's easy to navigate.

The document processing engine for specific types of documents, for the few docs that I have worked with it, is okay. It tries to mimic exactly what a human would do, but I wouldn't rate it at 100 percent. It does exactly what you want it to do but it might take a lot of time to get through a particular document. WorkFusion goes cell-by-cell compared to the same thing in Python where I could quickly just read it into memory and then do an entire page. With WorkFusion you can also read something into memory if you give it the right range and then you can use it in the way you want to.

As a bank, document processing is critical. You can't get it wrong. That's the reason that, for most of the bots we develop, we usually have manual intervention at the end for a human to review things. We get somebody to consent to it before it goes out because these are the kinds of things that commit the bank. We need to be extra careful with documentation.

We have started a center of excellence where we train people who want to learn how to do automation on their own. They might not necessarily be tech-minded, but we show them how to do it and then they can. We are getting a lot of people who are trying to get on board and buy into the idea.

A lot of people who are using it in our organization are using the free version. The certified version is something you pay for. We have about eight people who are using the certified version and we have two more people undergoing the certification process. The last time I checked, there were 20 users using the free version and practicing using the tool in general.

Initially, people thought it was geared toward somebody losing his job. But now they have bought into it. These bots are things that need to be handled and maintained by somebody. So it's not here to kick you out of your job but to provide you with an alternative route.

We have been in two proofs of concept with the solution. We're wrapping up two use cases with it right now.

The first one is for what we call "adverse media." We're required to do adverse media searches on potential new clients, as well during periodic reviews of existing clients.

For example, if we're onboarding a major retailer as a client, that retailer will have numerous negative news items, or adverse media. There might be employment abuse claims, civil lawsuits, and criminal investigations. For a large retailer, we would expect that kind of media. Currently, our adverse media process is performed by human beings, my analysts. They have to make a determination about the news items, based on rules set forth about negative news. If an analyst determines a match between an article and a rule, we would then escalate it to the business. The business would then make a determination as to whether the negative news is of enough concern that we would not onboard the client. Or, if the retailer was already a client, the business would determine how we should proceed.

With WorkFusion, their model will now use the criteria my human analysts follow, data elements to focus on first to validate whether the news is associated directly with our client. We first have to determine if an article is a true match. There are thresholds within the rules, severity levels. There's a difference between someone accusing the client of money laundering and whether they've been found guilty. Other severity levels are things like has someone been arrested, indicted, convicted, gone to jail, or served jail time. The Workfusion model is applying all those steps and thresholds that my analysts use, and should come to the same conclusion. The goal is that my analysts will be able to focus on the true matches and work on those. The point is to use our limited resources to focus on the high-value targets.

In our other use case, clients bring us documentation that we require to onboard them, such as articles of incorporation and ownership documents. Currently, what happens is that a banker would transcribe the key information from those documents into our onboarding system to start onboarding the client. After the customer has been onboarded, our review team will conduct a quality check against that profile to make sure the documentation is fully aligned with what was entered into the onboarding application.

We're using Workfusion's OCR model to read the documents that are collected and check the information in them against the profile that was created by the banker to determine if everything is correct. If it's not correct, it will then alert my analyst and tell them: "You now need to do a review and identify areas where there is a discrepancy to be corrected."

Once we close out our PoCs we will move the solution to Azure.

Before WorkFusion, my processes were all manual. Now, the Workfusion model reduces the amount of time my analysts have to spend reviewing client profiles that have been created against the documentation from the client.

We're in the process of assessing the results of the document processing engine. We have gotten very good quality or success ratios. They are in line with our expectations. The document processing engine is very important for our organization because, from our perspective as a bank providing financial services, documentation is the first step for onboarding any client. It touches everything we do, from lending to opening bank accounts to doing other transactions. Unfortunately, we're still in a paper-driven environment. It's critical that this be successful to reduce our handling time. It will bring greater efficiency and reduced costs. It should reduce our processing time by about 25 percent.

We have seen Workfusion improve on accuracy and compliance. As with any manual or human-driven process, you have to build a QC process. The QC process I have is manned by humans, my analysts. That's a cost, a redundancy we have to build. And it creates a "four-eye" approach to make sure that the person checking the work is meeting the procedure standards. Are they making errors that would cause us not to be in line with regulatory standards or expectations? With the Workfusion model, we reduce human error and the need for further QC. We switch more to a QA process where we're testing whether the model is effective.

The benefit to our organization will really be the overall efficiency. It's reducing the human processing time, which reduces costs and increases our speed to market. The sooner we can complete a review of a customer, the sooner the customer can start transacting and being fully engaged with our institution. Although I'm the head of our call center, anything I do to improve the revenue-generating opportunity of the business is a success factor I'm measured by.

I'm a non-technical person. I'm the business head. From what I've seen and been exposed to, Workfusion is very simple to use.

And in terms of the solution's flexibility towards evolving goals, it's highly flexible. During our PoC engagement, I have proposed functionality or enhancements and they've been open to them. They see value in the things that I have been proposing and that they would apply to their other clients as well. They have taken my suggestions for further development as future enhancements.

Where I think there is room for improvement is in the areas where I have suggested enhancements. There is enhanced functionality I would like to see, things that would further enhance the user experience, deliver greater efficiencies, and further mitigate risk.

We engaged WorkFusion in October of last year, so we've been using it for about five months.

My IT partners have raised no concerns about the stability of the product at all. There haven't been any hiccups in performance or any downtime during our engagement.

My impression is that Workfusion is very scalable. The purpose of my using it first in our organization was to test it out before using it enterprise-wide. I have sufficient volume and activity to prove out their model. And so far, the results are very promising. We're in the next phase of exploring enterprise-wide deployment.

My organization has about 72 people, a combination of offshore and onshore, and they are currently the users of the WorkFusion product. 

My partners in IT have had no issues with the support provided by WorkFusion.

Positive

We did not have a previous solution.

The initial setup was straightforward. Any hurdles that we encountered were the result of internal issues in our firm, and not because of Workfusion's requirements. With the way our IT infrastructure is built, it was a little bit more of a challenge for us. We're a smaller firm and we tend not to have the sophistication or the resources of larger clients.

We started the deployment in October and we started running in January. It took about 11 weeks to implement it, but that was more due to our internal structure.

Everything was done with the vendor.

We're going through the ROI analysis at the moment, but I can tell you that it's favorable. So far, they have hit all the success metrics.

The solution was not the cheapest, but it was also not the most expensive. Workfusion was very open to negotiating pricing and to working within the objectives of our firm.

We evaluated about four other vendors over a year ago through an RFP process that was managed by our procurement team. It took us about 12 months to settle on WorkFusion.

The reasons for not choosing the others included price or the viability of their products.

For someone who is looking into WorkFusion but already has an RPA tool in place, I would say that WorkFusion is a bit different. Workfusion's structure and strategy are not simply to be a single solution provider, but a fully integrated solution provider. They bring a whole suite of services to you. Others, like UiPath or Foxtrot Robotic Process Automation, just do one thing, RPA. But Workfusion brings all the services together, making it a one-stop-shop experience. Instead of your having to subcontract things out, WorkFusion's solution is all-encompassing. 

That was one of the key factors in our going with them because, in our next phase of development, they are going to be able to deliver that part, rather than our having to go to another vendor to complete things.

The biggest lesson I've learned is that when you're picking a firm, it's truly vital to assess how well they're going to work with you and how flexible they are. My past experience with other firms is that they're too rigid. They work within the confines of the terms and agreement and there's no room for flexibility unless we negotiate a change request or new terms.

With WorkFusion, what I have found is that there are firms out there that are willing to work with you, versus always sticking to the black and white of the contract you signed.

We have done a number of different workflow automations and process automations, spanning a large number of different business units. We have created bots with RPA. 

We have done a little machine learning and AI work with the solution, but not a ton. I haven't found a lot of use cases for this. 

It has delivered pretty good value for RPA and workflow automation, but we haven't gotten as much out of the machine learning, data extraction, and classification stuff.

The automations that we built are massively important to us. We have seen significant improvements. 

Over the past few years since we have built this out, it has permeated the organization. As we go and meet more people, they see what we have done or what we are doing and it has changed the way people think. For example, I have a number of business areas where they will tap us almost constantly, saying, "Hey. I have another good idea for automation." We have business leaders, and even people on the floor who are hands-on, throwing ideas out there and trying to make things even more efficient.

It has definitely helped us improve our accuracy and compliance. We don't use it a ton in the highest risk areas, but we are using it in a financial area and a number of different solutions. 

We have some time-sensitive transactions that are very difficult. You either have to hire an army or automate. So, we have automated some of that stuff. For example, if you need a PO change made, then you might not be able to get it all done in a day. If you are doing it manually, then some stuff might get recorded the next day. Whereas, with automation, you don't have that problem anymore. Things are more accurate partly because we don't get errors. We can build validation into our solutions and part of it is more accurate simply because it is more timely.

The ability to stitch together distinct pieces of a process into a larger workflow is the most valuable feature. I could emulate grabbing input from a screen, from an application, database, or Excel or a CSV file, then take that information and do a ton of different manipulations to it using rules and code, as necessary. Then, I can touch multiple other systems and grab more data or enter data into them, whether it be databases or front-end applications. That flexibility is huge in enabling us to really make a good impact with our automation.

The document processing engine is capable. We have a small number of invoices. For the most part, it works pretty well to process a document, extract the information, and then make updates in our accounting software. Invoices by their nature are structured documents. We only use it for certain vendors, so we are not just throwing everything at it.

It is not bad for structured documents. We just don't use it enough to be able to cheerlead for it. However, where we have used the document processing engine for structured documents, it seems to do a good job. 

WorkFusion does a lot of different things. In terms of workspace and document tagging, it is not bad and pretty intuitive. With a little bit of training, people generally pick it up and are off to the races. Other than that, we don't generally let non-technical users touch it.

We had an issue where they had a subproduct called RPA Express. We rolled it out and called it a power user program, which is basically your citizen developers, business people, and people who build smaller bots for their areas. We chose the RPA Express version of WorkFusion as the platform for which these guys would build their bots. We built up a program and got a bunch of people trained, then WorkFusion discontinued the product. I am not so sure that they always give our goals consideration, even though we are supposed to be one of their better customers.

I would like it if they provided a more detailed roadmap in terms of where they are going with the product. As we are trying to make decisions about the things that we are going to do to try to augment our capabilities, we could know whether they are thinking about the same things or not.

I have been using it just shy of five years.

It is moderately scalable. The scalability in version 9.0 is both a benefit and a drawback. 

The big problem with version 9.0 is if it is not specifically RPA, then everything that is not RPA (in terms of processing) happens on the application server, which is also where the orchestration happens. So, if you have data-intensive processes that aren't being distributed, since RPA and machine learning are the only distributed processes, then you have a major single point of failure. We have had some significant issues with this piece of it.

The newest version solves these drawbacks. In version 10.0, those tasks and the orchestration, which is the application server and the processing of non-RPA tasks, will be separated. The non-RPA tasks will have their own distributed clusters that you can scale, which is a huge difference.

We are looking to upgrade this year. The newest version (10.0) has more scalability for back-end processing.

They upgrade their product very often, but they don't offer enough time when supporting their existing product. I installed a new version of WorkFusion less than two years ago. They are now telling me it will be end-of-life this year. When you deal with most other enterprise software, the support period is like five to 10 years long. It puts a lot of strain on my entire group. We need to build up different environments and get all the infrastructure. The upgrade process is a nightmare.

The overall support could be much better. For example, when they say, "We are only going to support it for three years before you have to upgrade again," that is insane.

If I have something wrong with my environment and we create a ticket for WorkFusion, they get back to us pretty quick and are pretty helpful. They have certainly stepped up this piece lately. I would rate the technical support as eight out of 10.

Positive

The setup is extremely complex. Each time that we have upgraded, it has been extremely complex. This is part of my gripe. I am now doing my third upgrade in five years.

I came in during the initial deployment. That took a few months.

One upgrade took a year to do, from the first server being built to the last bot being migrated. This was because we had to migrate all our existing bots. This is the same process and time frame that we are expecting towards the upcoming upgrade. I will end up spending half a million dollars to do an upgrade.

We have seen ROI from using this solution. We saw value out of the solution very quickly. We saw it within a year. 

It really does help by either freeing up people's time to focus on more value-added tasks or reducing headcount. We generally have other things that we can get people to do, helping make the employees' work more meaningful since they are not having to do the mundane stuff.

The document processing engine is decent, and certainly not bad when we have used it. We find it to be capable. The biggest issue is just finding the use cases. I don't know if that is the vendor, us, or a combination thereof. I know when they have brought ideas forward to us, we are like, "We already have solutions." So, we have solved some of those problems already in other ways.

If the document processing engine broke tomorrow, we would not lose a ton of sleep over it. It might upset a handful of people, but they will live.

The digital worker is like a prebuilt solution that will replicate the steps or some piece of what a person may do today. We don't use any of that stuff out-of-the-box. Now, it is early days for them to be pushing their digital workers solution, but I have seen some stuff in their newer versions that might be of use. Part of it is the software capability, but the bigger part of it is we just don't need a lot of that stuff. 

One of the big ones that they always push is invoice processing, "We can scan your invoices and obtain the information off of them, then automate." It is like, "Okay, cool, but we get almost all our invoices through EDI, so we don't need that. It doesn't do anything for us." 

I would rate this solution as seven out of 10.

We use WF on banking use cases

It has helped us create new opportunities. 

WorkFusion has the capability to implement complex workflows, which is useful. ML capabilities are very powerfull as well

Non-code and further ML capabilities would be usefull.

Platform is very scalable with the exception of Control Tower. Some improvements could be usefull on this component.

A spanish customer service would be appreciated. I would rate the technical support as 9 out of 10.

Positive

WorkFusion's implementation is very easy. The biggest issue is analyzing the documents internally.

The implementation time depends on the project and use case.

WorkFusion is leader on IDP.

I would recommend WorkFusion as a very flexible tool.