If you are working out how to choose AI use cases, start the way Jason McKay does: define the problem, define the outcome, confirm the data exists, and keep testing until the person who asked for the tool says it makes a difference. McKay is the CIO of Riggs Distler, a company that upgrades infrastructure for utilities, and he picked PeopleReign for the work his team can’t build on its own.
McKay is candid that “a construction company is not a leading-edge type of entity,” so adoption depends on user experience. His technology team has sat in a general foreman’s transmission truck on a right-of-way to see what field crews actually need.
How do CIOs choose the right AI use cases?
The right AI use cases are narrow, measurable and backed by data that already exists. McKay begins every request by asking what problem needs solving and what outcome the requester expects. He then builds something fit for purpose and tests it relentlessly, so the person who needs it sees the value.
That discipline matters. Gartner predicted in June 2025 that over 40% of agentic AI projects will be canceled by the end of 2027, citing escalating costs, unclear business value or inadequate risk controls. McKay’s filter goes straight at those causes:
- Name the problem and the outcome. “I want to know the problem to solve, and I want to know what the outcome should be.”
- Check that the data exists. He asks whether data exists for the problem before anything gets built.
- Go narrow. Take a “very narrow AI approach,” building something fit for purpose and fit for use, then test it.
- Let the user confirm the value. The test is when the people who asked for it tell his team it is making a difference.
- Keep it simple. His rule for the rollout is “crawl, walk, run.”
Why does an honest AI vendor matter?
An honest AI vendor tells you when a use case isn’t a fit, which saves time and builds trust. McKay says PeopleReign did exactly that. After looking at a scenario his team was struggling with, PeopleReign said it wasn’t a great use case instead of selling him something that couldn’t work.
“We’re very deliberate in our approach. We’re not, quote-unquote, trying to AI everything. There are things that we know we can do, and there are things we know we just don’t have the horsepower to do. And that’s where PeopleReign comes in.”
McKay says that honesty told him a lot about the company. The use cases that did fit are getting over the finish line, with a focus on onboarding, data integrity and giving back-office teams what they need to finish payroll faster.
Where does Microsoft Copilot stop and PeopleReign start?
According to McKay, a general-purpose assistant like Microsoft Copilot can handle most of a company’s AI needs, and a specialized platform covers what it can’t. PeopleReign is built for IT and HR employee service and layers on top of existing systems rather than replacing them.
“Microsoft Copilot, I still argue at this time, can solve 80 to 90% of a problem that a company may have. And when it can’t, that’s when you go to a PeopleReign to help you with the things you can’t do.”
That is McKay’s own view, not a benchmark. It fits how he runs AI at Riggs Distler: Microsoft Copilot Studio is going out to project managers with AI lunch-and-learns every other week, and models are sized to the task. “We don’t need to be on the frontier,” he says. Gartner’s guidance points the same way: agents for decisions, automation for routine workflows, assistants for simple retrieval.
What does responsible AI look like at a construction company?
Responsible AI at Riggs Distler means keeping proprietary and personal data inside controlled environments, training employees, matching model size to the task, and verifying every answer. McKay’s rule for his team is that it is not just trust and verify. It is trust, verify and validate.
One example from his team’s own agent work: utility construction specifications can run to a thousand pages. Riggs Distler put an agent on top of them so crews can ask specific questions, such as what goes on a utility pole. Employees always ask for the page reference, flip to it, and confirm the answer, so they get speed without skipping the check.
What is Riggs Distler aiming for with PeopleReign?
Riggs Distler is using PeopleReign to help its teams with onboarding, data integrity and the back-office information that speeds up payroll. The goal is to turn work that took hours into minutes. McKay describes the rollout as crawl, walk, run.
The deeper goal is the employee experience: work that doesn’t feel like drudgery, and tools employees are excited to use. In his words, the most rewarding moments come from “seeing people who actually perform the work” cut lead times from two weeks to a couple of hours and gain insights they couldn’t get before. He also wants employees to become more editor than author.
Frequently asked questions
What is a good first AI use case for a CIO?
A good first AI use case has a defined problem, a measurable outcome and data that already exists. Riggs Distler CIO Jason McKay starts there, then builds something narrow and tests it until the requester confirms it makes a difference. Gartner similarly advises pursuing agentic AI only where it delivers clear value or return on investment.
Do enterprises need frontier AI models?
Not necessarily. McKay says most of what Riggs Distler does does not need heavy context and reasoning, so his team matches the model to the task instead of defaulting to the largest one. Part of his reasoning is cost: he wants to be responsible with tokens, the units AI models use to process text.
How do you build trust in AI with a skeptical CFO?
McKay shows economies of scale on real work. His team tested agents that summarize 20 completed projects at once, a job that would otherwise be done one project at a time, and that forecast headcount by trade three years out from the project pipeline. He also frames AI as a tool belt, not a fad.
Does PeopleReign replace Microsoft Copilot?
No. PeopleReign is an AI platform for IT and HR employee service that layers on top of existing systems rather than replacing them. McKay describes using a general-purpose assistant like Microsoft Copilot for most needs, and turning to PeopleReign for the things it can’t do.
What is PeopleReign?
PeopleReign is a people-first AI platform purpose-built for IT and HR employee service. Its action-taking AI helps employees resolve issues instantly, and it is designed to get a company’s main use case running in 30 days. It integrates with more than 100 enterprise systems, including ServiceNow, Salesforce, Zendesk, Workday, SAP SuccessFactors, Slack, Microsoft Teams and SharePoint, and works alongside them rather than replacing them.
Put AI to work where it creates real value
See PeopleReign in action and talk through the IT and HR use cases that matter most to your organization.
