Dr. V. Shravan Kumar
Technical Manager & ML Educator
Ph.D. Mechanical Engineering | M.S. Data Analytics - Georgia Tech M.S. Management - UIUC | M.Tech and B.Tech - IIT Madras
Engineering leader with 15+ years building high-stakes electronic control systems, predictive models and leading global teams at a Fortune 500 - now driving decisions through applied machine learning and building the curriculum to teach it.
About Me
A quick introduction
With a Ph.D. in Mechanical Engineering and over 15 years of industrial experience in the USA, my career has been defined by complex systems design and technical leadership. I’ve spearheaded global initiatives to architect model-based control systems, predictive models, and onboard diagnostics for mission-critical engine platforms while leading and mentoring teams in engineering data analysis, simulation-driven development, software-controlled hardware optimization, and technical problem solving.
Driven to expand my leadership impact and harness the power of large-scale engineering data, I completed master's degrees in Management (University of Illinois Urbana-Champaign) and Data Analytics (Georgia Tech). This dual foundation in executive strategy and machine learning enables me to bridge hardware, software, and advanced analytics.
These days, I split my energy between directing engineering teams and developing applied data science systems, ranging from time series forecasting, predictive modeling, and anomaly detection to ML-powered algorithmic trading and automation. I've also designed a graduate level course, Machine Learning with Python for Business Analytics, which I'm working to bring into an MBA classroom - because I enjoy teaching this technology just as much as I enjoy building it.
Projects
Data Science & Machine Learning Projects
WheelBot: Finance-AI-Automation
Algorithmic trading bot that generates income from options premium via the wheel strategy, selling cash-secured puts, managing assignment, then selling covered calls in a repeatable cycle with >40% annualized return. Combines sentiment analysis with a custom strategy and a monitoring dashboard.
Surplus-to-Invest Automation
Connects checking/credit card transactions (Plaid API) and a brokerage account, using ML to auto-categorize spending and forecast monthly cash flow. Detected surplus is recommended for transfer into an optimized portfolio, which also flags rebalancing trades when holdings drift from target weights.
Anomaly Detection
Benchmarked Regression, Random Forest, XGBoost, and LSTM models on transient time-series sensor data, with XGBoost winning on accuracy (R² 0.97) and speed for real-time deployment. Built a CUSUM-based anomaly detector that flags sustained residual drift, with SHAP analysis to explain feature impact.
Hierarchical Risk Parity Portfolio Optimizer
Built a smarter way to divide an investment portfolio across 12 stocks in three sectors. Instead of the classic Modern Portfolio Theory method, which gets unstable and often bets almost everything on just one or two stocks, it delivers a better risk-adjusted return (Sharpe ratio 1.49 vs. 1.05) while keeping a meaningful stake in every stock instead of an all-or-nothing bet.
Benford's Law Journal-Entry Fraud Screening
An audit prioritization tool for limited investigator time: Built a two-layer accounting fraud screener pairing Benford's Law digit-distribution testing with Isolation Forest anomaly model on journal-entry data. Flagging just the top 2% of entries by anomaly score caught 58% of injected fraud patterns (round numbers, threshold-avoidance, duplicate entries), a ~29x lift over random review.
Divvy Bike-Share Analytics
Analyzed a year of bike-share trip data (~780K+ rides) to uncover seasonal and weekday-vs-weekend usage gaps between casual and member riders. Built a Random Forest classifier to predict rider type from trip features (65% accuracy, ROC-AUC 0.70). Ride duration alone drove over half the model's predictive power, surfacing insights that could inform targeted marketing to convert casual riders into annual members.
Airline Network Optimization
Built a mixed-integer linear program (PuLP) to schedule flights across a 9-city, 25-route network with two aircraft types (narrow and wide body), optimizing flights per route, time-slot, gates, aircraft type, and pilot-crew constraints to maximize daily profit. Extended it with a Random Forest demand forecaster (R² = 0.84) that replaced hand-picked route limits, lifting optimal profit from $273K to $313K/day.
Credit Application Risk Classifier
Benchmarked 7 classification models — XGBoost, Random Forest, Neural Networks, SVM, KNN, Naive Bayes, and Logistic Regression on 20,000+ loan applications, with XGBoost reaching highest accuracy of 91.8%. Optimized the decision threshold with a cost-benefit framework balancing missed defaults against rejected creditworthy applicants.
Budget-Optimized Nutrition
Formulated a linear program (PuLP) that maximizes nutritional adequacy while minimizing cost across 10 food items, balancing calories, macronutrients, fiber, vitamin C, and iron requirements alongside food-group balance rules. Added a Random Forest model (R² = 0.90) predicting personalized calorie needs from age, weight, and activity level, then re-solving per person.
Experience
Where I'm working
Aug 2010 – Present
Columbus, Indiana, USA
Cummins Inc.
Controls, Software and Electronics Engineering – Technical Manager
Management & Leadership Experience
- Team Leadership & Delivery: Managed teams of up to 24 engineers, group leads, and contractors to deliver air-handling system controllers, predictive engine models, optimized electromechanical devices, onboard diagnostics, and data analysis for diesel and spark-ignition engine programs.
- Program Supervision: Supervised multiple engine development programs across on-highway and off-highway applications by managing deadlines and prioritizing tasks.
- Problem Solving: Addressed complex technical roadblocks and challenges using a structured seven-step problem-solving methodology.
- Global Talent Coaching: Coached and mentored international calibration and validation engineering resources in China, Brazil, India, and the UK to maintain technology leadership.
- International On-Site Leadership: Provided on-site engineering support at the East Asia R&D Center in China and Cummins-Dongfeng joint venture research facilities.
- Product Development Coordination: Coordinated Simulation-Based Product Development and next-generation engine architecture evaluations in global R&D facilities.
- AUTOSAR Control Systems Development Leadership: Led cross-functional teams in developing Cummins' AUTOSAR air-handling controller from stakeholder requirements through embedded software delivery and testing.
- Process & Standard Development: Restructured technical documentation, requirements, and engineering practices to integrate processes into global engineering workflows.
- Resource Management & Hiring: Partnered with HR, university recruiters, and consultancy firms to hire employees, interns, and contractors to fill critical technical gaps.
- Stakeholder & Team Alignment: Fostered collaborative environments, coordinated stakeholder voices, and successfully onboarded engineers onto clear career growth paths.
Technical Contributions & Engineering Experience
- Model-Based Controller Calibration: Engineered and calibrated model-based air-handling controllers across engine platforms to improve fuel economy, transient performance, and emissions compliance.
- Control Systems, Virtual Sensors and Diagnostics: Developed, calibrated, and analyzed model-based controllers, virtual sensors, electromechanical devices, and diagnostic algorithms to detect system-level failures.
- Data Acquisition & Vehicle Testing: Gathered and analyzed real-time performance and diagnostic data from test cell environments and prototype vehicles using data acquisition systems.
- Data Analytics & Visualization: Developed custom MATLAB data analytics tools to visualize key performance, OBD, and emission cycle metrics for stakeholders.
- Engine & Controller Simulations: Conducted engine simulations to design and evaluate new control system concepts for advanced air-handling architectures using GT-POWER and Simulink.
- Thermal & Fluid Systems Analysis: Applied core engineering principles to analyze intake air systems, EGR systems, heat exchangers, turbomachinery, combustion, and emissions.
- Embedded Software & Testing: Designed embedded software components, created prototype ECM builds, and validated software through open-loop/closed-loop bench testing.
- Onboard Diagnostic (OBD) Design: Designed a model-based OBD algorithm for air-handling subsystems to detect component failures on low-cost engine platforms (Six Sigma project).
- System Protection Algorithms: Developed a model-based turbocharger protection feature to prevent engine failure and ensure system reliability (Six Sigma project).
- Simulation-Driven Optimization: Integrated simulation tools with data analytics to evaluate model capabilities, refine virtual sensor accuracy, and validate electromechanical hardware performance.
Teaching
Sharing what I've learned
Machine Learning with Python for Business Analytics
Course designed for graduate/MBA students
A graduate-level course I designed covering the full ML toolkit for business analytics — supervised and unsupervised learning, gradient boosting, forecasting, and anomaly detection, with hands-on labs in Python (pandas, numpy, seaborn, scikit-learn, XGBoost, Prophet, statsmodels) and a group project applying it to a real business problem. Built on 15+ years mentoring global technical teams and translating complex analytics for non-technical stakeholders, skills I'm now looking to bring directly into an MBA classroom.
Graduate Engineering Courses (Instructor/TA)
Missouri University of Science and Technology
Developed, taught, and graded graduate-level engineering courses while pursuing my Ph.D., including a General Motors PACE (Partners for the Advancement of Collaborative Engineering Education) sponsored project integrating engine simulations (GT-POWER) into the mechanical engineering curriculum for Combustion Processes and Applied Thermodynamics courses.
Education
Academic background
M.S. in Data Analytics
Georgia Institute of Technology · Atlanta, GA, USA
Specialization: Business Data Analytics & Machine Learning
Aug 2023 – July 2025
M.S. in Management
Gies College of Business, University of Illinois Urbana-Champaign · IL, USA
Specialization: Finance and Data Analytics
Aug 2022 – July 2023
Ph.D. in Mechanical Engineering
Missouri University of Science and Technology · Rolla, MO, USA
Aug 2006 – July 2010
M.Tech, Energy Technology & B.Tech, Mechanical Engineering
Indian Institute of Technology Madras · Chennai, TN, India
Thesis: "Modeling of a Diesel Engine for Speed Control" · Minor: Industrial Engineering
Aug 2001 – May 2006
Skills
Tools I work with
Machine Learning & Analytics
ML & Data Libraries
Finance & Quantitative Analytics
Leadership and Management Skills
Languages & Tools
Engineering & Simulation
Achievements
Awards, certifications & recognition
Business Impact Award
Cummins Inc. — awarded by VP & Chief Technical Officer
2021 - 2022
Recognized for leadership and contributions to developing an AutoSAR air handling controller.
Systems Engineering
Cummins Inc. & University of Detroit - Mercy
Aug 2018
Completed a four-module systems engineering program: Innovation & Creativity, Product Planning & Design, Systems Architecture, and Systems Engineering.
Six Sigma Green Belt Certification
Cummins Inc.
July 2015
Certified Six Sigma Green Belt.
Business Analytics Certification
University of Illinois - Urbana Champaign
2022
Outstanding Under-35 Young Scientists Committee
HySyDays 2007, Second World Congress of Young Scientists on Hydrogen Energy
2007
Selected as one of thirteen members from eight nations. Organized by the Inter-University Research Center for Sustainable Development, Sapienza University of Rome, Italy.
US DOT National University Transportation Center Assistantship
US Department of Transportation - Research and Innovative Technology Administration
2006 - 2007
Competitive research assistantship supporting doctoral research on hydrogen for transportation applications.
Setting up Missouri's First Hydrogen Fueling Station and Operating Hydrogen Powered Bus
US Department of Transportation and Missouri University of Science and Technology
2006 - 2007
Research project to develop, demonstrate, evaluate, and promote safe use of hydrogen-based technologies.
6 Published Research Papers
Energy, IJHE, ASME IMECE and Professional Safety Journals
2008 - 2011
Published six international journal and conference papers from doctoral research on engine modeling, hydrogen combustion and safety; reviewed 30+ peer-reviewed papers for international journals.
Let's connect
Whether you want to collaborate on a Machine Learning project, or interested to bring the ML course into your classroom, I'd love to hear from you.
