Skip to content
Vignesh Winner

1

Headshot of Vignesh Winner

Headshot

Vignesh
Winner

B.S. COMPUTATIONAL ENGINEERING · B.S. MATHEMATICS · UT AUSTIN

Download résumé

2

Act I

Training

THE UNIVERSITY OF TEXAS AT AUSTIN

Degrees

  • B.S. Computational Engineering (Honors)
  • B.S. Mathematics — Statistics & Probability Pathway
  • Minor in Business

Austin, TX

GPA
3.88/4.0
SAT
1570/1600
ACT
35/36

Syllabus

  • Probability/Statistics
  • Stochastic Processes
  • Predictive Analytics
  • Game Theory
  • Time Series/Regression
  • Linear Algebra/Matrix Theory
  • Multivariable Calculus
  • Real Analysis
  • Macro/Microeconomics
  • Computational Programming
  • Software Engineering/Design
  • Engineering Computation
  • Scientific Computation
  • Scientific Engineering
  • Computational Fluid Dynamics
  • Computational Methods for Structural Analysis

3

Act II

The Roles

DRAMATIS PERSONAE — IN ORDER OF APPEARANCE

Role 06

Wrapped

May 2026 — August 2026

BP — SUPPLY, TRADING & SHIPPING

Quantitative Analyst Intern

  • Developed 5 Python-based OLS and time-series fair value models across multiple low-carbon products, translating regulatory, supply-demand, compliance, and fuel-market fundamentals into pricing signals that supported $5M+ in trading decisions.
  • Built a Power BI dashboard used by 20+ traders and analysts, centralizing fair values, model decompositions, coefficients, and market drivers while saving 2–3 hours of analysis per week.
  • Backtested systematic trading strategies using model-derived pricing signals across 10+ years of historical data, achieving 27% annualized return and a 1.3 Sharpe ratio under out-of-sample and regime-based validation.
  • Engineered end-to-end Dataiku pipelines for data ingestion and feature engineering through model validation and production deployment via Jenkins.
  • Gained cross-commodity exposure to crude and refined products through strategy reviews, trading simulations, and real-world case exercises.

StackDataiku · Power BI · OLS · Jenkins · Backtesting

New in this role + Dataiku + Power BI + OLS + Jenkins + Backtesting

Role 05

Now playing

August 2025 — Present

TEXAS MARKETRICS

Quantitative Finance Director

  • Designed quant curriculum spanning Python, probability/statistics, derivatives, risk modeling, backtesting, attribution, and regime stress-testing.
  • Built a 6-phase pipeline combining ARIMA forecasting, Markowitz optimization via CVXPY, and 10,000-path Monte Carlo simulations; stress-tested under −5%/−20%/−40% shocks with VaR/CVaR and backtested across 4 equities reporting Sharpe, Sortino, Calmar, and max drawdown.

StackARIMA · CVXPY · Monte Carlo · VaR/CVaR

New in this role + ARIMA + CVXPY + Monte Carlo + VaR/CVaR

Role 04

Wrapped

May 2025 — August 2025

GRADESOAR

Software Engineering Intern

  • Architected and shipped 6 full-stack production features in Next.js and React, including reusable question-bank and course-creation systems used by thousands of students and instructors.
  • Optimized RESTful APIs, MongoDB schemas, and data-access patterns across assignment and curriculum workflows, reducing retrieval latency by approximately 24%.
  • Strengthened platform reliability through backend routing, validation, and automated testing, increasing test coverage by 80% and reducing production defects by 15%.

StackNext.js · React · MongoDB · REST

New in this role + Next.js + React + MongoDB + REST

Role 03

Now playing

January 2025 — Present

TEXAS ENERGY TRADING

Quantitative Trading Lead

  • Built an N-BEATS forecasting model using NeuralForecast, pandas, and EIA API pipelines to predict WTI crude oil futures 3 months ahead.
  • Designed a multi-leg arbitrage strategy for a case competition using WTI/WTS spreads, contango futures, and pipeline congestion to optimize margins.
  • Led analyst training, case competitions, and trading-firm events while developing curriculum in derivatives pricing, energy fundamentals, and machine-learning applications to trading.

StackN-BEATS · NeuralForecast · EIA API · Arbitrage

New in this role + N-BEATS + NeuralForecast + EIA API + Arbitrage

Role 02

Wrapped

August 2024 — May 2026

TEXAS SPACECRAFT LABORATORIES

Flight Dynamics Lead

  • Led CubeSat mission analysis in ANSYS STK and Python across imaging, communications, sensor placement, and constraints within orbital configurations.
  • Conducted orbital trade studies comparing 500 km 28.5° and Sun-Synchronous (97°) profiles, finding 82% greater cumulative annual downlink time for the 28.5° orbit and identifying a −105° YPR star tracker mounting configuration achieving 0% obscuration.
  • Owned GNSS and ADCS subsystem testing, including OEM7600 receiver validation, pointing budgets, deorbit simulations, and interference analysis.

StackANSYS STK · Python · GNSS · ADCS

New in this role + ANSYS STK + Python + GNSS + ADCS

Role 01

Wrapped

August 2023 — May 2025

LONGHORN RACING SOLAR

Aerodynamics Engineer — Wheel Cover Lead

  • Led design and simulation of aerodynamic wheel covers using SolidWorks and ANSYS CFD, resulting in a 12% drag reduction in TACC simulations.
  • Fabricated structural components, including carbon-fiber aeroshell panels, bulkheads, and a dual-sided canopy latch to improve safety and egress speed.

StackSolidWorks · ANSYS CFD · Composites

New in this role + SolidWorks + ANSYS CFD + Composites

4

Act III

The Projects

RESEARCH

Fall 2025

Source

Weather-Signal Commodity Forecasting

  • Engineered 100+ features from NASA POWER weather anomalies, EIA inventory deltas, and price data to forecast weekly commodity futures returns.
  • Led a team of 4; trained XGBoost (Optuna-tuned) on 498-fold walk-forward CV across 5 futures — Brent, Henry Hub, ERCOT, Copper, Corn — reaching 57.8% out-of-sample directional accuracy with confidence-scaled LONG/SHORT/FLAT signals under a magnitude-threshold policy.
  • Built vectorized backtesting with transaction costs, slippage, and volatility regimes, yielding 13% CAGR and positive rolling Sharpe over 3+ years.

XGBoost · Optuna · Walk-Forward CV · NASA POWER

Spring 2026

Source

ML-Driven ETF Signal Research System

  • Engineered cross-sectional and temporal features across 25 ETFs, improving signal stability under walk-forward validation and purged CV evaluation.
  • Identified and corrected lookahead bias inflating Sharpe to 16; validated signals across 15 ETFs with a median Sharpe of +1.04 on holdout folds.
  • Applied Half-Kelly sizing and Benjamini-Hochberg correction across 250 model/ticker combinations with IC decay and SHAP feature diagnostics.

Alpha Research · Purged CV · Half-Kelly · SHAP

Summer 2025

Source

Tax Deductibility ML Classifier

  • Built and deployed a Flask-based ML pipeline using pandas and scikit-learn to classify tax-deductible purchase transactions with 89% accuracy.
  • Engineered a TF-IDF feature extraction process and trained a logistic regression model with confidence scoring and reasoning logic for classifications.

Flask · scikit-learn · TF-IDF

Spring 2025

Source

Space Launch Analytics API

  • Built a containerized Flask API with Redis to analyze 4,300+ Kaggle-based space launches via REST endpoints for timelines, maps, and sector trends.
  • Developed a modular Kubernetes-orchestrated job queue and caching system with Redis workers for async plot generation and scalable streaming.
  • Automated ingestion and analysis of launch data via Python requests, revealing a 1995 crossover to private missions using crossover and frequency plots.

Kubernetes · Redis · REST · Flask

THE PART THAT EXPLAINS THE RESTSummer 2025

Rehearse.AI

  • Developed an LLM-powered web app that analyzes uploaded scripts, infers character intent, and generates adaptive scene prompts for actor feedback.
  • Built a real-time speech analysis pipeline to evaluate tone, pitch, pacing, and emotional delivery using voice-processing libraries.
  • Engineered backend infrastructure using Flask, Redis, and Docker Compose for low-latency inference and production-ready deployment.

LLM · Flask · Redis · Docker

5

Act IV

The Craft

AFTER HOURS

Languages

Python · C/C++ · Java · JavaScript · MATLAB · SQL · R · Ruby · HTML/CSS

Web & Backend

React · Next.js · Node.js · Flask · MongoDB · Redis · REST APIs

Infrastructure

Docker · Docker Compose · Kubernetes · Jenkins · Linux · Git · TACC

Quant & Data

Dataiku · Bloomberg · Kpler · Enverus · Power BI

Simulation & CAD

ANSYS · OpenFOAM · Abaqus · SOLIDWORKS · Simulink · Ansys STK

Credentials

Columbia logo

Columbia Financial Engineering & Risk Management

Columbia

Dataiku logo

Dataiku Advanced Designer

Dataiku

Dataiku logo

Dataiku Core Designer

Dataiku

IBM logo

IBM Data Science

IBM

Google logo

Google Data Analytics

Google

Awards

  • Eagle Scout
  • Survivor Texas (S7) Winner
  • 2nd Degree Black Belt
  • National Qualifier — Duet Scene & Solo Monologue Acting

Off the clock

Theatre/Acting · Speedcubing · Ultimate Frisbee · Biking to Alaska · Derivatives Trading · Poker · Codenames · Charlotte/Kansas City Sports · Theta Tau · Delta Epsilon Psi