How One HR Team Put AI to Work to Improve Performance Management

Setting and enforcing employee goals can be one of the most tedious parts of performance management, especially when HR teams spend hours of time manually sifting through hundreds of individual goal statements. Yet when performance management is weak, it has the potential to undermine attempts to improve organizational agility, which was identified as a priority for 78% of CEOs according to SHRM’s 2026 CEO Priorities and Perspectives report.

Across the 16 HRX dimensions — from recruitment and compliance to talent management and leadership development — AI is increasingly augmenting the HR function. To demonstrate how HR professionals are putting AI into practice, SHRM reached out to members to share their experiences. This installment of our AI Use Case series explores how one HR team uses AI for quality checking of goals and objectives. 

A Virtual Workforce, A Manual Bottleneck

MedeAnalytics, a healthcare performance improvement company, builds advanced analytics solutions that transform data into actionable insights. According to chief people officer Lisa King, the HR team recently harnessed AI was to improve performance management.

Every year, after performance reviews, King's team would pull a full export of employee goals out of ADP, the company's Human Resource Information System (HRIS) — a spreadsheet covering the entire workforce. The objective was to confirm that every goal on record was a true SMART goal (specific, measurable, achievable, relevant, and time-bound) rather than a vague development goal such as "take a class."

That distinction was important because goal attainment feeds directly into the company's bonus plan. Previously, employees in a bonus-eligible role were paid out simply because the company hit its overall numbers, regardless of whether their individual goals were meaningful or even measurable, which King said wasn’t fair for those who were exceeding expectations. 

King and her entire HR team would also read every goal statement individually. Aside from the immense volume of statements to sift through, employees outside the U.S., some of whom weren't familiar with the SMART goal framework, needed extra explanation of what separated a SMART goal from a general aspiration. 

In need of a smoother process, King and her team turned to AI. Here’s how they streamlined performance management by quality checking of goals and objectives.

Sample Worksheet: SMART Goals Made Simple

The AI Use Case: Quality Checking of Goals and Objectives 

King's team now runs the same ADP goal export through ChatGPT Enterprise instead of dividing it up for manual review. The raw export includes around 20 fields per goal — including employee position, manager, and department data alongside goal titles, descriptions, start and end dates, and last-updated timestamps — spanning roughly 1,000 goals company-wide in a given cycle.

The process runs on two prompts. The first prompt normalizes the raw ADP data into a consistent format the second prompt can work from. Each of the company's divisional executives set their own management by objectives (MBOs), then check their employees' goals against those MBOs with the help of the second prompt, which has the same criteria for each goal. 

The second prompt is specifically built to do more than a simple pass/fail check. It applies what King describes as a “custom semantic alignment framework” — comparing employee goals to executive MBOs through business-outcome reasoning rather than keyword matching. 

Each goal gets sorted into one of three buckets: It directly owns an outcome, it contributes to that outcome through enablement or operational support, or it isn't aligned at all. For every goal, the tool generates an individualized rationale explaining why it landed where it did, then quality-audits its own rationale before producing a final evaluation in Excel.

Rather than flagging performance insights in general terms, the tool sorts employees by name and goal, King explained.

"[In one performance review cycle], it actually said, 'All these people are okay, these people are not. These people have two goals that are aligned, but these two don't align to the MBOs' — and we did that within ten minutes,” she said.

Because the two-prompt structure is reusable, King's team doesn't rebuild it each cycle — the same "normalization-then-evaluation" method gets run again the next time goals need auditing. Where the alignment call is still ambiguous, particularly for more technical, specialized goals, the team doesn't let AI make the final call. Those cases get kicked back to the relevant executive to classify directly.

The underlying approach of feeding structured HR data to an enterprise AI tool through a purpose-built prompt is one she's carried into other data-heavy corners of her job, including preparing workforce summaries for department leaders.

Cerification: People Manager Qualification (PMQ)

What Worked

The clearest gain was speed without sacrificing rigor. Instead of a team of people spending a workday or more combing through a spreadsheet, the review now happens almost immediately, freeing King's team to focus on coaching managers whose teams' goals needed rework rather than on the manual detection work itself.

It also made the process more consistent. Because the same prompt evaluates every goal against the same criteria, the review doesn't vary by which team member happened to read a particular employee's entry.

Specialty Credential: AI+HI

What Didn't Work

King was candid that the tool's usefulness depends entirely on the person reviewing its output — and that this is where she thinks a lot of HR teams could get into trouble.

"Someone who's never been in HR could look at this and say, 'Oh, that looks correct, I guess I've got to go with it,'" she said. "To be able to validate that it's real or not takes someone who already knows the subject matter, not just someone who knows how to run the prompt.”

According to King, AI lets her apply HR judgment to more people and faster, rather than replacing human judgment. 

The cultural and language gap didn't disappear either. AI could flag a goal as poorly formed, but King's team still had to have the conversation with international employees about what a SMART goal is and why it's different from a development goal — work that stayed human.

Human Oversight and Governance

Because MedeAnalytics handles sensitive employee and health-related data, King said the company requires staff to use its enterprise-licensed AI tools — Microsoft Copilot and ChatGPT Enterprise — rather than free or personal accounts.

The stakes are immense: The company is high-trust certified and handles personally identifiable information (PII) and health data as part of its core business, so AI tools needs to be secure.

Advice for HR Teams

King's advice for HR professionals looking to apply AI to their own goal-setting or performance data:

• Start with data you already have structured somewhere. King didn't build a new system — she deployed an AI tool to use the same HRIS data  her team was already pulling every year.
• Write the prompt from your own expertise. The tool is only as useful as the criteria you give it. King's team built prompts around the company's actual corporate objectives, not a generic goal-quality checklist.
• Insist on enterprise-grade tools for employee data. Free or personal AI accounts aren't appropriate for HR data, particularly when health or other sensitive information is involved.
• Keep a human in the loop who knows what "good" looks like. Someone with real HR judgment still has to confirm if a goal flagged by AI is correct while ambiguous, technical cases still get routed to the relevant executive rather than left to the AI's call.
• Tie the effort to something concrete. For King's team, the payoff was saved time and a fair bonus process.

"It really has transformed our work," King said of the shift. What used to take her team days of manual review now happens in minutes, giving them more time for the part of the job that still requires a person: The conversation with a manager or employee about what a good goal actually looks like.