Advancing Healthcare Through Optimization & Data Science
Healthcare Research Collaborator | Optimization & ML
I develop ILP/MIP optimization models for healthcare scheduling, codify clinical policies into reproducible constraints, and apply machine learning to broader research problems with measurable impact.
Seeking full-time healthcare research roles
Research Assistant · Health Informatics · Health OR · Clinical Analytics
Healthcare-focused research collaborator bridging optimization, ML, and clinical operations.
Healthcare Research Collaborator
I'm a research-driven analyst applying quantitative methods to healthcare operations and scheduling. I currently collaborate with a university faculty member on a healthcare scheduling optimization study, codifying clinical policies into ILP/MIP constraint families and contributing toward publication.
My broader work spans applied ML, simulation, and OR — from RL control to large-scale BI systems — brought back to healthcare problems where rigor, fairness, and reproducibility matter most.
Health InformaticsHealthcare ORClinical AnalyticsHealth AI
Education
MS, Business Analytics & AI
The University of Texas at Dallas
Expected May 2027 • GPA 3.7/4.0
Coursework: ML, Big Data Analytics, Statistical Modeling.
Research collaboration on optimization and scheduling.
B.Tech, Electronics & Communication
IIIT Nagpur
Graduated 2022
Walking-robot RL simulation in MATLAB/Simulink — 95% accuracy.
Foundations in programming, signal processing, mathematical modeling.
13+
Modeling Assumptions
Codified for healthcare scheduling
5
Constraint Families
ILP/MIP models for publication
20+
Papers Synthesized
Literature review depth
95%
RL Model Accuracy
Walking robot, 5,000+ iterations
40%
Simulation Speedup
Automated DL pipeline
20%
Capacity Optimization
Peak-hour metro operations
98%
Data Alignment
Weather-flight pipeline
30%
Faster Reporting
BI dashboard automation
Toolbox
Technical Skills
A breadth of tools across analytics, ML, and operations research.
Healthcare scheduling research alongside broader ML and optimization work.
Healthcare Scheduling Optimization (Research)
Faculty-led academic research translating clinical scheduling policies into ILP/MIP optimization models with cost, coverage, and fairness KPIs.
Problem
Hospital units face complex staff and resource scheduling problems where coverage, rest rules, and fairness must be balanced against cost — but policies are rarely codified in a way that supports reproducible optimization research.
Solution
Codified 13+ modeling assumptions and an evaluation plan, established a public-data acquisition workflow with a QA'd variable dictionary, and translated 16 policy items into 5 ILP/MIP constraint families (coverage, rest/consecutive, assignment, fairness, soft-coverage); synthesized 20+ papers to position the work for publication.
Impact
13+ modeling assumptions codified for reproducible research
AI Nutrition Companion: Real-Time Restaurant Meal Recommendation
Streamlit-based AI nutrition assistant that proactively guides users to healthier meal choices when eating at restaurants, using personal profile, daily intake, and menu data.
Problem
People trying to follow a healthy diet struggle when eating out — most calorie trackers only log meals after eating and do not proactively guide users before they order at a restaurant.
Solution
Built and deployed a Streamlit web app that estimates daily nutrition targets from demographics, activity, and goals, analyzes remaining calories/protein/fiber/sodium against meals already eaten, and ranks restaurant menu items with a rule-based engine classifying options as Light, Medium, or Heavy with clear explanations.
Healthcare AI project applying preprocessing and segmentation workflows to identify tumor regions in medical imaging data.
Problem
Manual review of medical images can be time-intensive, and tumor-region localization requires consistent preprocessing and segmentation support for clearer analysis.
Solution
Applied image preprocessing and segmentation techniques to isolate suspected tumor regions in brain scan imagery, strengthening practical exposure to healthcare AI, biomedical image analysis, and statistical evaluation workflows.
Impact
Built healthcare-focused image segmentation workflow for tumor-region identification
Applied preprocessing steps to improve scan consistency before analysis
Strengthened medical imaging, healthcare AI, and statistical analysis experience
Positioned as a foundation for future research publication
Walking robot simulation using MATLAB and Simulink, leveraging Reinforcement Learning to optimize control systems.
Problem
Complex locomotion control requiring intelligent optimization of robot movements across multiple trajectory patterns.
Solution
Developed walking robot simulation in MATLAB/Simulink with Reinforcement Learning for control optimization, and engineered an automated deep learning pipeline for neural network training and testing.
Impact
95% accuracy after training over 5,000+ iterations
40% reduction in simulation time through pipeline automation
4 movement trajectories mastered (straight line, circle, square, rectangle)
Optimized control systems for enhanced robot performance
MATLABSimulinkReinforcement LearningDeep LearningNeural NetworksControl Systems
Data-driven optimization system for Delhi Metro operations using Python analytics techniques.
Problem
Inefficient metro routes, unbalanced capacity utilization, and extended wait times affecting passenger experience across 200+ stops.
Solution
Analyzed 7+ datasets (stops, trips, routes, schedules) using Pandas and NumPy, developed spatial-temporal algorithms for stop density and trip intervals, and visualized hub density with Matplotlib and Seaborn.
Impact
20% boost in peak-hour capacity (6-10 AM, 4-8 PM)
10% reduction in off-peak trips optimizing resource utilization
15% reduction in average passenger wait times
10% improvement in overall metro system efficiency
Comprehensive analysis of flight delays correlated with weather patterns to identify key delay drivers and predict disruptions.
Problem
Airlines facing significant operational and financial losses due to weather-related flight delays without predictive insights.
Solution
Integrated 240+ GSOD weather station CSVs with 500K+ FAA flight performance records using Python (Pandas, Seaborn), automating data ingestion and transformation workflows.
Impact
98% accurate date-station alignment in weather-flight mapping pipeline
Quantified +7.5 min delay impact under heavy rain conditions
22% higher seasonal delay spikes identified in winter months
Top 10 reliable airports with 30% lower average delay vs. national mean
SQL-driven business intelligence project analyzing movie rental operations to optimize inventory, customer engagement, and revenue.
Problem
Movie rental business lacking data-driven insights for inventory management, customer segmentation, and stakeholder engagement.
Solution
Delivered comprehensive reports by joining 16+ tables mapping store managers, inventory, and stakeholders across 10+ locations with advanced SQL aggregation of 1,000+ films and transactions.
Impact
Enhanced store-level planning visibility across 10+ locations
Identified rating-based inventory distributions and top-value customers
12% increase in repeat customer retention through improved segmentation
Streamlined board engagement by linking transactions, advisors, and awards
Machine learning model to predict loan defaults for the Small Business Administration loan program.
Problem
High default rates impacting SBA loan program efficiency requiring accurate risk assessment and real-time scoring capabilities.
Solution
Engineered 15+ analytical features (loan-to-asset ratio, processing efficiency), trained an XGBoost classifier optimized for AUCPR on imbalanced data with early stopping and threshold tuning, and deployed via Gradio on Hugging Face Spaces.
Impact
AUCPR of 0.472 achieved on imbalanced loan data
Real-time loan scoring capability via Hugging Face deployment
Enhanced model transparency using SHAP and permutation importance
Reusable scoring pipeline for stakeholder insights
Collaborated with senior engineers on code reviews and scalable data systems.
PythonAlgorithmsProblem Solving
Published on LinkedIn
Newsletter & Insights
Sharing knowledge on data science, analytics, and business intelligence.
Professional Insights Featured
By Mohan Venkata Pavan Sai Teja Kattiboyina
Professional Development Dec 2023· 4 min read
Creativity Is Valuable — But Execution Is Essential
Do's and Don'ts in a Professional Setting — Part 1. While creativity starts the conversation, it's execution that delivers results. An idea has value only when it's executable.
