Skip to main content
AIBlogBot SupportRPARPA and AI

Why are RPA and AI Managed Services necessary?

By August 23, 2023September 19th, 2023No Comments4 min read
Why are RPA and AI Managed Services Necessary?

It’s stated that approximately 50% of Intelligent Automation projects fail (RPA, AI, IDP). Some of the contributing factors are insufficient process analysis, poor license utilization, no coherent support model, varying methodology for building, inability to acquire certified resources, growing tech debt tied to resources, inability to keep up with the latest tech, and slow scale hindered by IT and internal resources. My company, CampTek Software, was founded on the premise of solving these problems from inception. We realized from the beginning that Managed Services for RPA and AI are necessary to build, support, maintain, and scale an automation program. Many have attempted to create an internal COE without guidance and assistance or are intent on building it themselves. While there are some that have done this successfully, most of them are Fortune 500 companies with the budgets and resources to sustain a successful COE. We have encountered even at that size; that managed services are needed. This gives some indication that it’s less about having a budget for COE resources and more about having seasoned experts who can support automation solutions.

Simply, why are RPA and AI Managed Services necessary?

Support… Support… Support…

Surprisingly (or not so surprisingly) the most overlooked and biggest driver to ensuring the long-range success of any automation program is…I attended Forward II over five years ago and the tagline was a “bot for every human.” I thought to myself, “That’s great, but who is going to support these bots? How are they going to work every day?” Organizations continue to overlook this essential idea. Hiring a staff of developers will not cure this. Most developers don’t want to support their own code, they want to create new code. There is a paradox we often encounter after so many bots are built by a developer who is trying to create and support simultaneously, there is a law of diminishing returns in that they spend more time supporting versus creating new bots so the program can start to cannibalize itself. Not to mention it leads to poor job satisfaction which can lead to developers leaving. Hopefully, in this case, the PDDs are up to date and the code was built to an industry standard. Otherwise, the bot’s maintenance can hinder its performance.

Hosted Support and License Utilization

One of the major complaints about RPA since the enterprise players started selling licenses to companies is the over sell and/or underutilization of licenses for the Unattended/Attended Bots. If the license is purchased and it isn’t used, it costs money. These costs can rise particularly for larger bot purchases. Managed services eliminate that problem in that the Service Provider will only charge what is needed based on the bots usage versus having to buy licenses ahead of time that may never get used.

Hosted Support is also a key driver to speed to scale and ensures increased reliability of the program. IT departments for mid-market companies are the most strapped for resources and are overburdened handling the internal needs of the company. The never-ending focus on security and keeping both contemporary and legacy systems properly maintained, in addition to onboarding/off-boarding users, are all blockers from helping scale a fully-fledged automation program. Bot machines need to be added or subtracted based on capacity in a timely fashion. Managed Services can make that happen and provide the same or better level of security and governance that an on-prem managed solution can. With the proper setup of nonpublic IPs and a secure private network, one can have the safety and control of on-premises architecture but with the luxury of being able to order virtual machines that can handle any need (RPA, AI, and/or IDP).

An internal COE can’t do it alone

Customers often indicate they want to “bring it in-house” and build their own internal COE so not to depend on an external company to do the work of building and supporting the automation in production. I think this is a flawed concept for small and medium-sized companies. We have found that every company, regardless of size, needs a partner to start, and most continue to need one to manage the maintenance and growth of their automation stack. The technology is rapidly getting more robust, and it’s virtually impossible for an internal COE to keep up with the latest tools and best practices. As mentioned previously, “support” is often overlooked. It astounds me, as this is the first thing I think of when building an automation program. Most developers don’t want to do the support, and if they are charged to do it, more time will be spent on supporting versus developing as the program grows. In essence, it’s short-term thinking that an internal COE can effectively build as well as support. My thought is that the best use of an internal COE is to concentrate on building the pipeline of automation opportunities, developing new automations, either through Citizen Development or traditional means, and providing governance around the program.

CampTek Software was one of the first companies to realize that managed services and RPA as a Service are the best path to ensure success and get the highest ROI out of the program quickly and efficiently. If you would like to see our approach to this, please review our overview here.

Written by: Peter Camp. CTO & Founder