Building Quant HR – from reporting to signal-driven decisions

Predicting collective human behaviour is hard: low signal-to-noise and shifting structures. This session applies a quant fund model – signals, back-tests, portfolios and attribution – to people decisions. We map a practical architecture, show the reporting scaffold, share a campus-hiring mini case, set realistic guardrails for GenAI, and close with a six-step replication playbook. Examples reflect previous implementations; the current direction is aspirational and in build.

• A CFO-friendly quant fund mental model for HR: expected value, risk and benchmarked deltas.
• Quant HR architecture: data management → data science → action → monitoring with human-in-the-loop.
• Reporting as scaffold: self-serve, automated feeds, daily regression tests and normalisation discipline, with a client analytics exemplar.
• Mini case: campus hiring signals, model evaluation, fairness controls, attribution and a human final gate.
• GenAI's real use today, policy as a temporary output and a six-step replication playbook.

Taras Kowaliw

Managing Director, Head of People Analytics · Blackstone

Senior leaders need faster, defensible workforce decisions under cost pressure, talent scarcity and rising model-risk expectations. Finance expects auditable attribution, clear benchmarks and measurable impact on cycle time, hit-rate and cost. HR needs fairness, bias controls and robust data foundations. A quant-style approach frames trade-offs, tests assumptions and enables repeatability.