It may come as no surprise that artificial intelligence is a foundational aspect of the Now Platform experience. If you watched any of the recent NBA or NHL playoffs, you almost certainly caught Idris Elba explaining what ServiceNow’s AI agents will do for you.
(Mr. Elba was unavailable for comment at the time of writing.)
There is real value in what Agentic AI brings to the table, and last month, we took a deeper dive into Agentic AI use cases. Of course, before getting into use cases, or maybe even just testing the AI waters, it’s good to understand what genuine AI readiness looks like.
To better understand how to prepare for and implement enterprise AI, we checked in with CoreX Technical Architect Aaron Munoz. Normally, after a subject matter expert interview, we would quote our expert in a standard narrative article format. But after speaking with Aaron, we quickly realized that his answers were simply too good to truncate.
So, here is the interview, complete and unaltered, for your enjoyment.
***
David Kirkpatrick: Where should companies start when getting into an AI deployment?
Aaron Munoz: There is a concept of “garbage in, garbage out” anytime an automation tool is introduced to enhance a process, and it’s no different with AI. Whether it’s data within the company systems or external connections, without accurate and up-to-date information, the solutions will not be trusted and will be abandoned.
Expectations for the technology and limitations need to be communicated with the user base, and security policies must ensure the company is compliant without sacrificing user experience. Having in-house expertise on AI and tool-agnostic architecture ensures independence from vendor lock-in and owning your own AI roadmap.
How about when approaching an AI deployment in ServiceNow?
Being up to date with the most recent release of ServiceNow allows companies to have the latest tools available. The solutions depend on the products that are installed (think ITSM, CSM, HR, and so on) and the licensing level (professional vs enterprise). Certain tools are ready to use from the day of go-live, like Virtual Agent conversations. But others require enough historical data to train models like Predictive Intelligence for categorization.
To maximize the value of AI in a ServiceNow implementation, having a phased approach and getting quick wins helps to get additional buy-in from the customer’s management and push for additional adoption. The original plan can be assessed against new business requirements.
With all of this in mind, is there a roadmap or checklist when thinking about AI readiness?
Focusing on value rather than available tools is the way to go. It’s easy for technical consultants to suggest tools customers aren’t ready for or expecting to use without an initial discovery of the problems to be solved. A large part of the effort for this is getting the correct people in the conversation from the get-go.
There is also a separate situation of customers asking for solutions that don’t have a well-defined problem to be solved, or even the correct license to implement. Having a proof-of-concept demo helps to ground the ideas into actionable requirements that can be discussed and estimated.
Is there a basic assessment list for AI readiness?
Assessments can be broken down into the following: business value, data quality, technical feasibility, adoption readiness, risk, and compliance.
What do companies need to know about the Now Platform’s AI capabilities?
Companies need to understand the value of having AI capabilities within the platform, ready to use with minimal technical knowledge needed to get started. The models are made specifically for the way ServiceNow works.
External tools are not required, but can also be leveraged from the most common AI asset providers. By using the existing flows and other built-in tools in an instance, AI agents can perform actions autonomously or with human supervision.
I’d like to get into AI and ServiceNow modules. Are there use cases in different industries to point to?
The modules offered by ServiceNow tend to have built-in solutions out of the box. They tend to cover the most common scenarios, but it’s up to the implementation team to tailor them according to customizations or build the solutions from scratch, similar to building for a custom application. The more digitized a process is, the easier it is to adopt AI to reduce manual effort.
Examples in manufacturing are predicting maintenance based on anomaly detection. For retail, it’s helping customer service personnel with the summarization of previous support cases. For industries like healthcare and telecom, there are concerns about data privacy and security that must be met.
What does stakeholder buy-in mean for a successful AI deployment?
Executive-level buy-in allows for breaking down silos and allocating the budget for the implementation effort. Business use cases for introducing AI solutions are not possible without the buy-in from the process owners. The end user who will use the AI solutions also needs to adopt the new way to work, which improves their productivity and makes work easier.
How should companies prepare and educate their workforces for AI?
It depends on the roles that will adopt the AI usage that is introduced. End users learn that AI is an additional option to trigger their requests and see the status of their prior submissions. Fulfillers receive tips for knowledge and similar solutions to current requests. Managers can get updates on how their processes are doing compared to expectations.
Something common in all roles is the importance of maintaining data privacy when using AI tools that are not officially supported by the company, plus understanding the limitations of the current AI models.
Okay, I’m ready to deploy AI at my organization. How do I choose the first use case?
First impressions are critical, so the first use case sets the tone for AI that can enhance the company’s processes. The use case should be high impact, frequently needed, quick to deploy, and relatively safe to implement and adopt.
I have my high-impact AI implementation in place. Now, is it better to start slowly and build out my AI strategy piece-by-piece, or should my plan include deploying AI across business units and workflows?
A phased approach, where a complete vertical slice is released, allows a quick return on investment and serves as a proof of concept without losing momentum with a prolonged development time. [But] having a roadmap to guide future efforts serves as a “north star” for the next phases.
One situation where releasing in multiple BUs and workflows works is a complete replacement of another AI solution with the ones supplied by ServiceNow.
I want to know how my AI deployment is performing. What metrics should I keep track of?
Most KPIs can be created within the platform using Platform Analytics for measuring the usage of Now Assist and Agentic AI in ServiceNow. Some KPIs that help bring visibility: Adoption Rate, Time Saved, Deflection Rate, Success Rate, and CSAT.
Setting a target and thresholds will also give context to the actual results by comparing them to organizational objectives.
What should anyone thinking about AI readiness understand?
Improvement can come by changing the process, the people, or the technology. While the focus here has been on the technological solutions ServiceNow offers as part of its platform, having a good organizational change management to support people in the adoption and standardize the process with best practices is still required for a successful implementation.
Instead of a perfect solution, continual improvement helps to get value quickly and with control from a functional and governance point of view.
What is the big takeaway on AI readiness?
All industries are already using AI to get ahead and leaving behind the competition that is lagging in adopting it. It may be overwhelming to even identify the starting point for this journey. Relying on a trusted partner that has walked this path and can provide guidance is the best approach to get ready with confidence.
***
As you can see, the interview was wide-ranging and covered everything from high-level concepts on AI readiness to tactical details on deployments. The bottom line is that AI is table stakes for any organization in today’s business world. Fortunately, ServiceNow is at the forefront, providing tools, guidance, technology, and more for companies ready to fully embrace the AI present and future.