Model validation, financial systems, and research judgment.
Professional work across volatility-model monitoring, FX settlement validation, compliance-risk testing, commodity pricing, financial-data controls, causal-inference research, and evidence-based AI evaluation.
Quantitative Finance Expert, AI Evaluation & Benchmark Design
Contract, Remote · Sep 2026–Present
Built quarterly model monitoring for a HAR volatility service against EWMA across 512 forecast origins and eight equities; scored 478 origins on RMSE, MAE, bias, and relative performance through a policy-version transition.
Constructed 21-session realized-volatility outcomes and isolated a variance-reported-as-volatility unit break in one of eight series despite a release note describing the change as service-wide.
Designed nine independent domain judgments; one frontier model failed all three runs, with a units error cascading into five wrong statuses and five unwarranted reversions. A 96-criterion rubric scored the gold-standard solution 96/96 versus 62/96, with AI graders calibrated over seven runs to match every assessment.
SUMMER ATLANTIC CAPITAL
Quantitative Engineer
Internship, Remote · Oct 2026–Present
Own four engineering work items and two model-risk documents—evaluation, monitoring and alerting; retraining, versioning and rollout—on a 60+ item build plan.
Built a Python/pytest validation suite for multi-currency FX settlement math: six independently computed scenarios, zero mismatches, and five open policy questions including rounding and inverted-rate risk.
Stress-test a weighted four-factor risk score at band boundaries and build API authentication, tenant isolation, and production guards; auditing also surfaced an unbounded identity-risk input and three backend APIs not yet connected.
02 / BRIGHT POWER · CURRENT ROLE
ENERGY MARKETS ANALYST, COMMODITY PRICING & FINANCIAL DATA ANALYTICS
Market pricing and financial controls across energy portfolios.
New York, NY · Sep 2026–Present
$4.91Msales-tax portfolio$1.16Moutstanding refunds$256Kfees reported47%of pending value in five cases
COMMODITY PRICING
Analyze electricity and natural-gas supplier bids, contract tenors, usage profiles, and utility benchmarks across Edge on Demand, EnergyScoreCards, and Salesforce; present recurring updates on NYISO and commodity-price drivers.
FINANCIAL RECONCILIATION
Lead 101 refund cases within a 179-claim, $4.91 million sales-tax portfolio, using SQL and Excel to reconcile filed claims, payments, and fees.
ROOT CAUSE & CONTROLS
Traced $81K in refunds and fees to wrong-batch assignments, incorrect submission amounts, and fee-rate discrepancies; specified 68 corrections by system, record, and field.
EXECUTIVE REPORTING
Report $1.16 million in outstanding refunds and $256K in fees to the VP, Commercial & Energy Markets; five cases hold 47% of $680K pending at NYS.
Constructed workflows extracting, normalizing, and validating utility data across provider portals, OCR-processed documents, and partner systems.
SQL RECONCILIATION
Wrote multi-table queries reconciling account, billing, building, and provider records across five operational systems.
CONTROL DESIGN
Specified a validation control that reduced manual review by approximately 50% while surfacing duplicate and misconfigured records.
OPERATING JUDGMENT
Resolved more than 390 data-integrity exceptions and earned a performance-based pay increase and nine-month extension.
04 / RESEARCH & EVALUATION
AHMED LAB · RUTGERS INSTITUTE FOR HEALTH
Lead Research Scientist & First Author, Applied Statistics
New Brunswick, NJ · May 2023–Apr 2024
First-authored a peer-reviewed Wiley publication recognized as a Top Viewed Article in 2025.
Evaluated instrumental-variable estimators and robustness methods across 150+ studies, including two-stage least squares, Egger regression, block jackknife resampling, and Bayesian and mixture models.
Directed an eight-member team through research scope, synthesis, and expert-reviewer revision.
Scored more than 100 artifacts against defined accuracy, completeness, contextual fit, and instruction-following criteria.
Cited specific evidence, identified unsupported claims, and delivered structured evaluations under time-bound production requirements.
05 / EXPERIMENTAL SYSTEMS
QINGZE ZOU LABORATORY · RUTGERS SCHOOL OF ENGINEERING
Research Assistant, C++ Data Acquisition
Piscataway, NJ · Oct 2023–May 2024
Reliable measurements begin before analysis.
Developed and debugged a C++ data-acquisition system for timestamped 1 Hz sensor logging. Systematic component testing isolated hardware-software integration faults and reduced experimental data loss by approximately 50%.