Punith Kumar

Data Preprocessing: Engineered Python pipelines (Pandas, NumPy) reducing data cleaning time by 35% and improving dataset quality for downstream ML tasks.

ML Modelling: Developed & evaluated 3+ classification models (Logistic Regression, Random Forest, SVM), improving predictive accuracy by up to 20% over baseline.

EDA: Performed exploratory data analysis on structured datasets; identified key patterns enhancing feature selection and boosting model performance by 15%.

Data Science Intern - Prinston Smart Engineers - Davanagere, India

(2026-01 - 2026-05)

• Data Preprocessing: Engineered Python pipelines (Pandas, NumPy) reducing data cleaning time by 35% and improving dataset quality for downstream ML tasks.
• ML Modelling: Developed & evaluated 3+ classification models (Logistic Regression, Random Forest, SVM), improving predictive accuracy by up to 20% over baseline.
• EDA: Performed exploratory data analysis on structured datasets; identified key patterns enhancing feature selection and boosting model performance by 15%.
• Reporting: Delivered detailed project reports and model evaluations supporting data-driven decision-making across the team.

Data Science Intern - Medical Cost Prediction Project - Prinston Smart Engineers - Davanagere, India

(2026-01 - 2026-05)

• ML Pipeline: Built prediction system for annual medical costs using Random Forest & Linear Regression on health and demographic data.
• Model Performance: Random Forest achieved higher accuracy with lower RMSE by capturing non-linear relationships in insurance cost factors.
• Feature Engineering: Engineered features on age, BMI, smoking status and region, improving prediction reliability by ~18%.

B.E. - Artificial Intelligence & Machine Learning - GM Institute of Technology, Davanagere (2022-08 - 2026-05)

P.U.C. - Science (Higher Secondary) - Science Academy PU College, Karnataka (2020-06 - 2022-04)