 |
🚀 Data & ML practitioner with 3+ years of experience in analytics, machine learning research, and production ML pipelines. 🤖 I design ML systems that are reliable, observable, and deliver business impact — cloud-based pipelines, well-tested models for churn and risk, and dashboards that push insights to decision-makers. 📊 I work across Python, SQL, Docker/Podman, Apache Airflow, and cloud platforms, with a strong focus on experiment tracking and model evaluation for predictive analytics. 💡 Open to ML System Design, MLOps, and Cloud-Native Predictive Analytics roles. |
🟢 Machine Learning Specialist — Florida Gulf Coast University (Apr 2025 – Present)
- Architected end-to-end predictive analytics pipelines with Podman/Docker and Apache Airflow on 100k+ historical records
- Tuned CatBoost + Optuna classification models achieving 80%+ accuracy and 10–15% performance gains for churn/risk prediction
- Integrated predictions into Oracle DB and reporting workflows, enabling self-serve dashboards for non-technical stakeholders
🟢 Graduate Research Assistant — ML & GIS — Carl Vinson Institute of Government, UGA (Aug 2023 – Dec 2024)
- Built deep learning models (Mask R-CNN, Res2UNet, SegFormer) for satellite building damage detection; peak AP of 0.8757
- Engineered end-to-end ML pipeline on A100-powered HPC cluster with Neptune experiment tracking
- Developed Leaflet-app: an interactive web tool for real-time ACLED conflict data visualization with GitLab CI/CD
🟢 Graduate Student Researcher — University of Georgia (Aug 2023 – Dec 2024)
- Led MS Research on multivariate time-series forecasting of traffic flow using 100+ California PeMS sensors
- Achieved LSTM top performance: MAE ~7.9, R² ~0.99 on traffic flow prediction
🟢 Graduate Teaching Assistant — University of Georgia (Aug 2022 – May 2023)
- Mentored 150+ students in Data Science and Discrete Mathematics; contributed to a 35% increase in high-quality project submissions
- Led project-based labs in Python, Java, and C++ with a 95% course-completion rate
🟢 Data Analyst (Intern) — Bro Services (Apr 2021 – Oct 2021)
- Built cloud-based ETL pipelines on AWS, cutting reporting latency by ~50% and improving forecasting accuracy by ~20%
- Designed Power BI dashboards with DAX/SQL/Python, driving ~25–35% improvements in conversion and BI adoption
| Degree | Institution | Details |
|---|---|---|
| MS, Computer Science | University of Georgia | Full tuition waiver via TA performance |
| BE, Computer Engineering | Savitribai Phule Pune University (AISSMS IOIT) | First class with Distinction & Honors in Data Science |
| Project | Description | Skills |
|---|---|---|
| Student Churn & Risk Prediction | CatBoost + Optuna pipelines on 100k+ records via Airflow/Docker at FGCU | MLOps, CatBoost, Airflow |
| Traffic Flow Forecasting | Multivariate LSTM on 100+ CA PeMS sensors (MAE ~7.9, R² ~0.99) | LSTM, Time Series, PeMS |
| Satellite Damage Detection | Mask R-CNN/SegFormer on Maxar imagery for 2023 Morocco earthquake (AP 0.8757) | Computer Vision, Deep Learning |
| World Population & Building Density Map | Interactive heatmaps using Mapbox & Turf.js | GIS, JavaScript, Mapbox |
| HR Analytics Dashboard | Power BI dashboard for employee attrition & salary analytics | Power BI, DAX, SQL |
