OmniCorp's Digital Transformation in Talent Acquisition

OmniCorp, a fast‑growing Singapore software company doubling its headcount every two years, is weighing whether to adopt HireMe, an HR analytics tool that scrapes public professional data and uses machine learning to score candidates on five pillars (Skill‑Match, Sociability, Conscientiousness, Leadership Potential, and Cultural Fit). The promise is to eliminate costly proactive sourcing and potentially replace early interview rounds. Using data on 450 previously interviewed candidates, students assess whether the scores predict hiring outcomes, whether those relationships are causal, and whether OmniCorp should adopt the tool, opening up discussion of prediction versus causation, selection bias, and the role of algorithmic tools in human hiring decisions.

By the end of the case discussion, students should be able to:
1. Distinguish prediction from causation, and articulate why a variable that predicts an outcome in observational data may not produce that outcome if acted upon.
2. Identify selection bias in a hiring dataset and reason about how it limits the inferences that can be drawn.
3. Evaluate the validity of algorithmic candidate scores when the “ground truth” being predicted is itself a human judgment that may encode bias or noise.
4. Apply a basic cost benefit framework to a build versus buy or adopt versus reject decision involving a vendor tool, incorporating both labour cost savings and risks.
5. Reason about where in a decision pipeline an algorithmic tool should sit, and the trade offs of each.
6. Surface ethical, legal, and fairness concerns specific to algorithmic hiring, including disparate impact, privacy of scraped data, opacity of proprietary models, and gaming.