|
I’m a student building end-to-end data science projects from real datasets. I enjoy the whole path: asking better questions, building a dependable baseline, evaluating honestly, and turning the result into an interactive demo. My portfolio explores machine learning, NLP, customer analytics, cybersecurity, experimentation, recommendation systems, and open-source developer tooling. |
|
01 Understand |
02 Model |
03 Deliver |
|---|---|---|
| Find patterns and assumptions in the data | Compare baselines and stronger models | Turn results into a useful dashboard |
ContribCheck: GitHub Issue Readiness Checker
An evidence-based Python CLI and FastAPI service that checks dependencies, assignments, competing pull requests, repository health, and CI before contribution work begins.
|
Hybrid Movie Recommendation System Content-based discovery, collaborative filtering, hybrid ranking, and offline evaluation. Mobile Game Retention Experiment A/B testing, retention analysis, bootstrap uncertainty, and product decisions. Telecom Customer Churn Prediction Churn risk, threshold-aware classification, retention economics, and segmentation. |
Network Intrusion Detection System Binary intrusion detection, attack diagnosis, explainability, and web-log monitoring. Six-category harmful-content classification across online text. Hotel Reviews Sentiment Analysis Classical ML, LSTM, transformer inference, and an interactive dashboard. |
Languages and tools used across projects and coursework; experience depth varies by tool.
Languages
Data and scientific computing
Machine learning, NLP, and vision
Visualization and applications
Workflow
I compare meaningful models, document trade-offs, preserve reproducible workflows, and keep the limitations visible.
