Md. Nawshin
Navin

• Engineered an enterprise-grade web application from scratch using C#, .NET, and JavaScript to master complex business logic, resulting in a production-ready system built strictly on OOP principles and MVC architecture.
• Collaborated in an agile-driven environment using Git version control and participating in code reviews to simulate real-world team workflows, which significantly reduced debugging cycles and ensured high code maintainability.
• Implemented industry-standard design patterns and secure APIs under the mentorship of senior engineers to bridge the gap between academic theory and commercial software development, proving the agility to rapidly adapt to enterprise tech stacks.

• Engineered end-to-end predictive pipelines using Scikit-Learn, Pandas, and NumPy to accurately forecast used car resale valuations based on multi-dimensional vehicle features, establishing total data integrity by treating outliers and structural anomalies.
• Applied advanced feature engineering including temporal age extraction and non-linear log transformations (np.log1p) to resolve highly skewed metrics, alongside robust ColumnTransformer layouts to safely automate categorical encoding without data leakage.
• Benchmarked ensemble architectures across five regression models, optimizing a final Random Forest Regressor to achieve a peak predictive accuracy of R² = 0.963 and an MSE of 0.817, encapsulated into a production-ready real-time inference function.

Led creative initiatives and coordinated team-based artistic projects. Managed event planning and execution within departmental responsibilities.