Ray Liao

Built a full driver-based financial model covering revenue build, OPEX structure, working-capital requirements, and contribution margin tree; integrated scenario & sensitivity tables to evaluate FX exposure, volume/price drivers, and cost-to-serve economics across 3 multinational portfolios Performed revenue bridge & margin bridge analysis using ~117,800 transaction-level financial records; decomposed variance into mix-shift, utilization efficiency, and input-cost effects, surfacing ~14.7% addressable cost variance and informing FY25 restructuring options Created executive forecasting dashboards (Power BI) that unified revenue cadence, gross-margin waterfall, expense buckets, and budget vs. actual variances; reduced FP&A reporting cycles by ~37% and standardized review metrics across 5 business units Consolidated financial assumptions (WACC inputs, volume ramps, productivity curves, overhead-allocation logic) across Finance, Operations, and Strategy teams, reducing cross-functional reconciliation time and strengthening forecast alignment

Built Python-based forecasting models (ARIMA + STL decomposition) on 21 years of demand & price data; integrated elasticity coefficients, raw-material price curves, and procurement lead-time assumptions to identify ~¥13.8M (~$1.9M) in sourcing efficiencies Conducted TAM/SAM/SOM market sizing and developed unit-economics and LTV/CAC mini-models, applying sensitivity analysis on churn rates, frequency curves, and discount factors to inform Manner Coffee’s FY24 expansion strategy Evaluated store-level P&L across 23 locations by decomposing COGS into input-price variance vs. efficiency variance and mapping labor productivity vs. capacity constraints; the resulting margin-driver model supported ~9.4% OPEX improvement Used root-cause analysis, OPEX-bridge modeling, and cost-benchmarking to recommend inventory-planning, waste-reduction, and supply-chain-cycle optimizations that improved scalability and reduced order friction

Analyzed customer lifecycle and revenue pipeline data from ~5,200 client interactions; built retention-cohort curves and a unit-economics model (LTV decomposition + servicing-cost allocation) that supported a ~19.7% uplift in repeat engagement Conducted segment-level profitability analysis by benchmarking CAC, servicing cost, contribution margin, and labor intensity; identified two client cohorts with negative contribution margins driven by high servicing cost and low upsell probability, leading the team to recommend reallocating account-management hours toward higher-value cohorts