Phat Ngo

Senior Technology Consultant · Atlanta, GA

Global SAP S/4HANA transformation, Fortune 50 pharmaceutical client

• Applied AI and anomaly-detection models to production monitoring workflows, flagging access anomalies and role misassignments to cut errors 65%; partnered with client risk stakeholders on complex control problems.
• Designed predictive KPI models and attribution reporting across a $487M program; identified workflow inefficiencies and automated reporting processes, cutting issue-resolution time 10+ days.
• Built Power BI dashboards and analytical reporting serving 300+ daily users globally; scaled best practices across teams, and insights contributed to $103M in program revenue.
• Presented model methodologies, assumptions, and performance results to senior executives, defending quantitative approaches under scrutiny.
• Drove firmwide adoption of AI and agent-based workflow automation across the SAP practice, improving how work is generated, reviewed, and executed; mentored 16 consultants on analytics tooling.

Personal, live capital

• Design, backtest, and run automated equity-index futures strategies in Python: regime filters, session-based risk limits, and simulation-first execution.
• Monitor performance, risk exposures, and attribution daily: Sharpe and Sortino ratios, maximum drawdown, drawdown duration, and out-of-sample stability, with hard loss limits enforced in code.
• Built a mean-variance portfolio optimizer in Python (Pandas, NumPy) applying modern portfolio theory to size and rebalance allocations across strategies subject to volatility and drawdown constraints.

Personal research, published on this site

• Built a driver-based DCF from SEC filings for Rivian (deliveries × price, separate software & services line, stock comp expensed) with bear, base and bull cases, a reverse DCF and a WACC-by-multiple sensitivity; rated Neutral with a $14 probability-weighted fair value.
• Initiated NVIDIA (Bullish, $240 fair value) and AMD, SpaceX and Even Realities as labeled conviction calls, each shown beside the model’s probability-weighted value ($313, $54 and $0.40B) on one shared DCF kit: AI revenue as gigawatts × revenue per gigawatt, warrant dilution, segment builds and reverse DCFs.
• Delivered an Excel model for every report (699 live formulas for Rivian) with a source note on every input, reconciled to the published web report to the cent.

Georgia Tech

• Developed and validated regression models on EIA, weather, and CAISO market data; ran residual diagnostics and benchmarked Elastic Net (R² = 0.945) against gradient boosting (R² = 0.964).

Georgia Tech

• Built an end-to-end NLP pipeline on earnings-announcement text: tuned LDA topic models, fit Lasso regression linking disclosure language to firm outcomes, and documented the model lifecycle.

M.S. in Analytics — all three tracks completed (Computational Data Analytics, Business Analytics, Analytical Tools)

Coursework: Deterministic Optimization · Machine Learning for Trading · Simulation · Probabilistic Models · Computational Data Analysis

B.B.A. in Computer Information Systems, Data Analytics & App Development

CFA Level I candidate (exam scheduled) · AWS Cloud Practitioner · Lean Six Sigma Green Belt · SAP Activate Project Manager · SAP Generative AI Developer