The deployment took twice as long as we originally scheduled.

The implementation of WorkFusion's machine learning models for OCR and intelligent document processing was supposed to take six months, but it ended up taking one year.

The deployment team included two primary developers, the CMO, and the digital worker product owner from WorkFusion. On our side, we had two developers, a PMO, and a group of business analysts to develop the requirements and conduct testing.

The implementation process began with a proof of concept where we reviewed our process at a high level and WorkFusion demonstrated how their digital worker, Daryl, could achieve our key goals for that process. From there, we moved into the full business requirements development phase, which took a couple of months. This was followed by the technical requirements phase, the development phase, and the system integration testing phase. Once the SIP tests were passed, the implementation went to user acceptance testing. Finally, there was a month of bug fixes before the implementation was pushed to production.

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. 

It is complex for enterprise implementation. The solution needs a needs minimum, of 20-22 servers. It's no small task. This is not a straightforward setup. It's a large undertaking and there are many moving parts.

Deployment takes at least a month. In terms of getting everything done, a company is looking at a deployment of 15 to 20 days minimum.

The initial setup was fine. WorkFusion has smart people on its team. It took between 21 days and one month.

The people involved in the setup on our side were not the technical people but, rather, the functional guys. They are the people who handle the documents on a day-to-day basis. We were supported in terms of the technical requirements, like how many users use an application and what the workflows were that should be done by the bots. A team of nine to 10 people, the functional and the sales and marketing people, participated.

There isn't much maintenance. It is well-maintained by WorkFusion's admin. You can check if there are any updates or upgrades available, depending on your licensing, but all that information is available.

I wasn't part of the very first dialogue with execution, however, it was quite good, actually. We got into contact with the senior leadership straight away and had a good dialogue, and we have had a good dialogue all through sort of it as journeys together with WorkFusion. That's one of their strengths of working with them. You get direct contact. And if needed, it's easy to escalate and get the support you need.

Getting everything up and running took quite a while. However, it was early in the company. It might have taken around a year or so. 

The initial setup has gotten simpler. It used to be a very complex setup, but WorkFusion has made a conscious effort in terms of simplifying the installation processes. The newer versions are a lot simpler compared to the previous ones.

It took too long to implement their machine learning models for OCR and Intelligent Document Processing. For our business process, it took little over two months to implement. However, we were not getting the desired level of accuracy. So, WorkFusion provided its own experts, then the team worked together for roughly six months.

We have had WorkFusion for a little over two years. We are still in the process of rolling out. We haven't been dynamically updating our existing business processes because of a change or something, but our team has been flexible and nimble to respond to the changed business environment requirements. That helps provide the capacity, but it is still a development cycle that takes a couple of months from conceptualization to deployment. That is where we are. It is definitely good to have the extra capacity where you can go and deploy it. Overall, conceptually, it is good to have capacity if an organization is facing a scenario where goals are changing frequently. Bots could definitely help. We haven't faced the situation where our goals change that frequently.

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.

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.

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.

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.

The initial setup was straightforward and a very good experience. The WorkFusion team worked seamlessly with our sanction screening vendor to turn it around quickly and easily from our hardware end.

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.

The initial setup was straightforward. We did not have too many issues, just with the number of servers and the cost of hosting it full-time. 

It took about three to four weeks to implement their machine learning models for OCR and Intelligent Document Processing. The main timeframe was with document tagging and the training data. Our SMEs, data engineers, data scientists, and data analysts were involved in implementing this functionality.

I was not involved in the initial installation as it happened before I arrived. I can only say that I have not heard anything about the issues that may have occurred.  

It does require maintenance and there is ongoing support from a technical training standpoint. Then the bots themselves require maintenance and updates. That is, if there is anything that changes in the environment or any process changes, the bots may need to be reconfigured — or even replaced — to work with the changes.  

Setup - Unable to comment

Support - Support has grown and continues to expand

The setup was time-consuming but simple. We had the skills and resources to complete it without a problem.

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

The implementation time depends on the project and use case.

Speaking about implementing the entire business process, for a medium-sized implementation that includes document recognition and the RPA platform, it took three people about a month and a half. That started with collecting the requirements and ended with the final demo and acceptance.

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.

I was hired specifically to support this environment.

When I came on board, it was already implemented, rolled out, and already in production.

I've been involved with implementing software upgrades for the product with vendor assistance.

The initial setup was pretty straightforward. That may have been because we had the guys who built the system from Minsk come and do it for us. The initial installation took six months and that included doing a proof of concept.

Setting up the server and the desktop versions are quite straightforward, both for the studio and for the control tower. As long as you have the required access, the setup process is straightforward. You just have to press the Next button on a series of screens.

We had a deployment on IIS and I built a web application on top of the bot. In total, the deployment took about a week because there were some functionalities that were not being picked up properly. Once those were completed, everything worked properly.

Post-setup, the deployment was a little bit difficult.

In terms of maintenance, one or two people are sufficient. You need a developer, and then somebody to support the infrastructure.

The initial setup is straightforward.

The initial setup was straightforward. Implemention took 13 to 14 weeks and configuration also took the same time, 13 to 14 weeks. We used the help of a consultant to implement. It was good working with them.

The person doing the installation of this solution needs to be someone that is skilled in this department in order to complete it. If you are new to the installation you will have difficulties.

The initial setup is straightforward because we also have a license agreement with WorkFusion. We are a reseller of WorkFusion as well.

If it's a simple process, then the deployment can take around two weeks from start to end. If it is a medium process, it takes three to four weeks. And any complex process might take five to six weeks.

The initial setup was very straightforward. 

The initial setup is not very complex. It's easy if you don't want too much of a data science tool. You do need to hire some data scientists if you want to be sure about regulations. If you have some regulators coming into the office to check, you'll want to be able to answer questions from them.

The installation and initial setup are okay. It is better than some RPA products but not as good as UiPath.