Framing before answering.
I take ambiguous business questions apart, isolate the real constraint, and rebuild the problem into something the team can act on. The answer is cheap once the question is right.
Chapter 02Work
Marketing analytics, AI systems, and the team rhythm behind better decisions.
My work sits across performance marketing, experimentation, forecasting, measurement strategy, and AI-enabled workflows. I'm usually brought in when the data is messy, the system is fragmented, and the business needs a decision it can actually trust.
How I work
Three habits behind the output.
Decisions, not dashboards.
Measurement work only matters if it changes a budget, a bid, or a plan. I build the bridge from noisy media signals to the specific decision leadership needs to make.
Products, not just analysis.
When the problem repeats, I like turning the workflow into a product. That can mean an internal AI agent, a forecasting system, a reusable SQL view, or a consumer app.
Experience
The career path in chapters.
Across media, higher education, and edtech, the through-line stays the same: understand behavior, structure the problem, and turn analysis into action.
Jun 2025 — Present
Marketing Analytics Manager
WPP Media
Lead global performance marketing analytics across U.S., Canada, and multilingual campaigns, translating live CPA, ROAS, and ROI signals into optimization actions, measurement plans, and faster spend decisions.
Built a Newton AI agent to surface in-flight optimization opportunities faster.
Scaled Luxe IQ across 5 to 6 accounts, reducing client-ready strategy output from weeks to minutes.
Manage 3 Senior Associates and mentor 10+ analysts across 3 lines of business.
Sep 2022 — Jun 2025
Senior Associate, Business Intelligence
WPP Media
Built analytics foundations for new accounts, shipped AI-enabled tools including Luxe IQ and CapPilot, and partnered with data engineering on SQL, BigQuery, Adverity, taxonomy QA, and dashboard migrations.
Developed CapPilot with 4 specialized XGBoost models, reducing normalized forecasting error by 73.5%.
Led a 90-day Datorama-to-Power BI migration with SQL, BigQuery, Adverity, ETL mapping, and automated taxonomy QA.
Automated MMM data collection from about 1 week to under 20 minutes for Circana and Nielsen workflows.
Oct 2020 — Nov 2021
Digital Marketing Analyst
Johns Hopkins University
Ran cohort and funnel analysis for acquisition campaigns, supporting 130% site-user growth and data-driven redesign decisions across departments.
Analyzed acquisition funnels, cohorts, and behavioral engagement to guide channel and content decisions.
Partnered with UX and creative teams on redesign and content changes adopted across 5 departments.
Feb 2020 — Oct 2020
Marketing Analyst
Baicizhan
Led app-install campaigns, A/B experiments, and influencer-driven growth work that improved conversion, active users, and commerce performance.
Improved mobile conversion by 25% and active users by 20% through campaign and onboarding optimization.
Supported launch and iteration of a community and short-form video feed using user feedback and engagement analysis.
Toolkit
The tools support the thinking.
I care less about listing every platform and more about using the right combination of methods, models, and communication to make the answer useful.
SQLPythonRBigQueryPower BITableauLooker StudioAdverityAPIsAttributionMMMIncrementality TestingXGBoostForecastingFeature EngineeringGoogle AnalyticsAdobe AnalyticsAI Workflow DesignPrompt Engineering
Want the full resume, or to talk about a role?