Data System Diagnostics
I examine how metrics are defined, how teams are incentivized, and where reporting systems may be distorting reality. I identify which metrics actually reflect real outcomes and which ones are misleading proxies that teams may be optimizing instead.

Trust Signal Modeling
I analyze operational data to identify hidden friction, KPI drift, and patterns like repeat contacts that signal underlying system problems.

Actionable Decision Translation
I translate statistical findings into concrete operational changes — process fixes, measurement improvements, and governance recommendations.

AI systems learn from the data organizations give them. When that data is shaped by gaming, incentive distortion, or measurement failures — the AI learns the wrong thing and accelerates the damage.

I build detection systems that identify where metrics have decoupled from reality, where training signals have been corrupted, and where AI is optimizing toward outcomes that look good on paper but harm the people it serves.

My research is grounded in a decade of direct observation inside one of the largest telecom retention operations in the US — and validated through NovaWireless, a synthetic AI lab built to model exactly these failure patterns at scale.

I treat data as a product of real systems—people, incentives, tools, and processes. Most problems aren’t ‘bad analysis’ — they’re bad definitions, broken incentives, or missing context.

Then I move quickly through three steps:
• Clarify the decision and what “success” actually means
• Validate the data (definitions, gaps, bias, incentive effects)
• Translate results into actions: process changes, experiments, or models

The goal is clarity you can act on—not another dashboard you ignore.

Tell me what decision you’re trying to make and what data you have. I’ll tell you what’s possible and what I’d recommend as a first step.

No sensitive data needed—IDs only are fine.