If I was going to give any advice to someone who was just about to utilize the software, I would say that the most successful organizations that apply this technology make it an entire team effort. It's not started in one business unit. And if it is, it's socialized across the enterprise. That's the quickest way to scale: getting everybody onboard. The second-biggest thing is that the most impactful projects you will get will come from your people, your internal workers. And until you get them to understand what the software can do and its capabilities, it's going to take you longer to scale your program. So make sure everybody's socialized, and make sure everybody truly understands what the software can truly do. They're going to give you the best opportunities to benefit from it.
Deployment should definitely be done using the support of experts. Even when I owned my own RPA program, and I wasn't in a consulting capacity, I still reached out to a third-party to get support. While setting it up is something that you can do internally, given that most objectives include speed-to-market and quick scaling - wanting to see results in 60 days instead of six months - it's going to be very difficult to do alone, especially if your goal is to have 100 bots in a year. If your goal is ten, you can probably manage it. It's important to use experts if you are looking to rapidly scale.
I have implemented UiPath in virtual environments, including on-prem, Azure VMs and servers, SQL-based data storage, as well as AWS. I've never had any issues with the responsiveness or the application having any problems operating. The biggest consideration that you have when trying to deploy robots in a virtual environment is making sure that your architecture is sound. You have to integrate through severs and you have to take into consideration firewall updates. And then there's interacting from the cloud if your applications are on-prem. You have to make sure that the bot doesn't have any issues. But if your architecture is solid and your infrastructure is set to support the applications in a cloud environment, there shouldn't be any issues. You wouldn't notice any difference compared to having them on a desktop on-premise.
I would agree that UiPath eliminates human error, but I would add the caveat that good code eliminates human error. I've been doing this for a while and I've seen bots that mess up. It's in your delivery methodology. If you have a sound delivery methodology - you're going through a rigorous UAT cycle and are having outputs audited by the subject matter experts - you should literally get to zero errors. Maybe you will have five percent exception cases, but your error percentage should be zero.
Having worked with all the tools, they all have little niche components. As long as UiPath continues to focus on knowing what the next wave of technology is that businesses really need to use to be efficient, and they start embedding that skillset in their software, that's all you could ask for. They need to stay in front of the power curve of technology, which is impossible, but they're trying.
I've never had a bad issue with UiPath. My experience with them has always been pleasant and engaging. They're never stuck at just giving you software, showing you how to use it, and then walking out. They're always focused on improving your business. If you focus on that, and focus on generating value, you can't lose.
Automation technology is the number-one driver across an organization now. Trying to find ways to do more with less has been the going mantra for organizations for years now. It's no longer feasible to simply run operational efficiency or Six Sigma projects to try to get gains. The only way that you're going to get significant gains is going with an automation-first approach. That's where I see a lot of organizations headed, even spending more on RPA software than on cloud implementation. It's a very big focus, and I don't see that slowing down any time soon.
On a scale from one to ten, I would you rate UiPath as an eleven. It's excellent software.