Reimagining Enterprise Performance Management with AI
Designing an AI strategy to help managers make better performance decisions—not just write better reviews
Role: Product Design Manager
Project: Concept exploration
Duration: 1 month
Team: 2 Product Designers, 1 UX Researcher
Overview
Annual performance reviews require managers to synthesize months of observations, goals, peer feedback, and business context into fair, evidence-based evaluations.
While generative AI is often viewed as a tool for writing, we saw a larger opportunity: using AI as an ongoing coaching partner that helps managers make better decisions throughout the entire performance cycle.
I led a design exploration that reimagined how AI could reduce administrative work, improve consistency, and increase manager confidence without replacing human judgment.
The Opportunity
The traditional performance review process is highly manual.
Managers must gather information from multiple sources, summarize months of employee performance, prepare calibration materials, and participate in review discussions—all while trying to ensure fairness and consistency across their teams.
Much of this effort involves organizing information rather than applying leadership judgment.
We believed AI could shift that balance.
Instead of helping managers write faster, it could help them think better.
My Role
Sponsored the concept work.
Helped define the product vision.
Facilitated design critiques and cross-functional reviews.
Worked with Product and Research to identify high-value opportunities for AI throughout the performance lifecycle.
Coached the design team as concepts evolved.
Helped communicate the long-term vision to leadership.
This was intentionally a future-looking initiative focused on identifying meaningful product opportunities rather than delivering production-ready features.
Rethinking AI's Role
Many AI experiences begin with a blank text box.
We intentionally avoided that approach.
Instead, we asked:
Where do managers actually struggle?
The answer wasn't writing.
It was decision-making.
Managers needed help understanding context, identifying gaps, preparing for conversations, and maintaining consistency over time.
That insight shaped the entire product direction.
The Vision
Rather than appearing once a year to draft performance reviews, AI would become an ongoing partner throughout the performance cycle.
Continuous context
Instead of starting every review from scratch, AI could continuously organize relevant information throughout the year, helping managers build a more complete picture of employee impact.
Guided review preparation
When it came time to prepare evaluations, AI could identify missing information, surface relevant accomplishments, and ask thoughtful follow-up questions that encouraged managers to reflect more deeply before finalizing their assessments.
The goal wasn't replacing manager judgment.
It was improving it.
Calibration support
One of the most complex parts of performance management is ensuring fairness across teams.
We explored how AI could help managers prepare for calibration discussions by highlighting inconsistencies, surfacing areas with limited evidence, and identifying where additional discussion might be valuable.
Rather than recommending ratings, AI would increase confidence by making reasoning more transparent.
Looking beyond the review
Performance conversations shouldn't end once ratings are submitted.
The vision extended into the following performance cycle by helping managers identify development opportunities, generate future goals, and support ongoing coaching conversations.
This shifted AI from being a document-generation tool into a year-round management assistant.
Takeaways
This project challenged the notion that the most valuable AI experiences simply automate existing tasks. They improve human decision-making.
By focusing on the moments where managers experience uncertainty—not just where they spend time—we uncovered opportunities to create a product that could strengthen leadership, improve fairness, and reduce administrative burden simultaneously.
For enterprise AI, that's where I believe the greatest opportunities exist: not replacing expertise, but augmenting it.
