Analyzed 150,000+ Uber ride bookings and built an end-to-end supervised classification pipeline in Python to predict booking outcomes using Logistic Regression, Random Forest, and XGBoost. Improved baseline performance by 22% through feature engineering, SMOTE-based class imbalance handling, and hyperparameter tuning, achieving 99.7% test accuracy with an optimized XGBoost model.
Sucheta Nandy
I build scalable machine learning systems and data pipelines that power decisions at scale. Currently pursuing an MSE in Data Science at the University of Pennsylvania.
About
I'm a Master's student in Data Science at the University of Pennsylvania, with a B.Tech in Computer Science & Engineering (2022) and three years of industry experience as a Systems Engineer at Tata Consultancy Services. That mix of production engineering and graduate-level ML research is what drives my work. I care just as much about a model's real-world reliability as its F1 score.
My sweet spot is the full pipeline: cleaning messy real-world data, building and evaluating classification and NLP models, and deploying them on cloud infrastructure (I hold GCP Professional Data Engineer and Associate Cloud Engineer certifications, plus Azure AZ-900). At Penn, I've gone deep on everything from logistic regression and decision trees built from scratch to neural networks in PyTorch, always with an eye on reproducibility and rigorous evaluation.
When I'm not tuning hyperparameters, you'll find me hunting for sneaker drops or nerding out over car specs - two hobbies that, like ML, reward patience and attention to detail.
View selected projectsSelected Projects
A genomic text processing pipeline using NLP to extract meaningful biological insights from large genomic datasets.
A responsive To do list app with local persistence and deploy pipeline to Netlify. Used for UX experiments and portfolio demos.
Technical Skills
- Programming Languages: Python, SQL, C, C++, Java
- ML & Data Science: PyTorch, scikit-learn, XGBoost, Pandas, NumPy, Matplotlib, Plotly
- NLP: spaCy, NLTK, Hugging Face Transformers
- Big Data & Streaming: Apache Kafka
- Cloud & DevOps: GCP (Professional Data Engineer, Associate Cloud Engineer), Azure (AZ-900), Git, Unix/Linux
- Databases: Oracle SQL, PostgreSQL
Soft Skills
- Cross-functional Communication: Translating technical findings into actionable insights for non-technical stakeholders.
- Analytical Problem-Solving: Breaking down ambiguous, open-ended problems into structured, testable approaches.
- Project Ownership: Driving projects end-to-end from scoping and prototyping through deployment and handoff.
- Collaboration & Mentoring: Working effectively in teams and helping peers ramp up on new tools and concepts.
- Learning Agility: Quickly picking up new frameworks, languages, and domain knowledge as projects demand.