What Should HR Actually Build With AI?

HR does not need more AI ideas. SHRM recently mapped 138 AI use cases across HR.¹ The harder question is deciding which ones are actually worth building.

One of the simplest ways to make that decision is the classic Impact–Effort Matrix.

Impact is the value an idea could create. Effort is the time, cost and complexity required to make it work.

Plot each idea in one of four quadrants: Quick Wins, Major Projects, Fill-Ins and Thankless Tasks.

A weak first draft is easy to correct. A bad employment decision is not. The higher the consequence of an error, the stronger the testing, review and oversight need to be.

Where most HR teams should start.

In most cases, the best place to begin is Quick Wins and a few low-risk Fill-Ins. They are easier to test, easier to improve and faster to prove.

A morning brief is a simple Fill-In. With an approved AI tool connected to calendar and email, it can summarize the day, prepare for meetings and surface messages that need attention.

It is not transformational. It does not need to be. If it removes recurring work at very low cost, it has already earned its place.

AI changes what is worth building.

A few years ago, even a modest internal tool could require a developer, budget approval and weeks of work. Many good ideas never cleared that hurdle.

AI changes the economics. Smaller ideas can now be tested faster and for far less money.

Instead of proving that an idea deserves a full development project, HR can build a useful first version, test it with real users and decide from evidence whether to keep investing.

One example is Wellhana’s interactive ROI calculator. It lets HR leaders adjust assumptions and immediately see how the potential business case changes.

The same logic applies across HR. A recruitment ROI calculator could sharpen a budget request. A leadership briefing builder could turn workforce data and source material into a clearer case for action. An employee-feedback analyzer could surface patterns that otherwise take hours to find.

None of these needs to become a giant transformation project. Build the smallest version capable of proving or disproving the idea.

Earn your way into the bigger projects.

Major Projects can create more value, but they demand more from the organization. Data has to be cleaner. Permissions matter. Integrations get harder. Testing gets stricter.

A natural-language HR analytics tool is a good example. The upside can be substantial: leaders ask workforce questions in plain English and get useful answers without waiting for another report.

But the tool is only as strong as the data, access controls and logic underneath it. That makes it a very different build from meeting preparation or first-pass feedback analysis.

Start with smaller, lower-risk work. Prove that AI can produce a useful result. Learn what breaks. Build internal confidence. Then take on the heavier projects with better information and a stronger case.