SUSHMA SRIRAMULA

Cleaned and modeled datasets using Excel and Power BI, creating dashboards for 5+ client scenarios.

Delivered data-driven insights through visualizations, improving decision-making by 15% in simulated client presentations.

Built data pipelines for 1,000+ records, ensuring data integrity and consistency.

Presented insights to a virtual team of 6, receiving 90% positive feedback for clarity and impact.

Built a command-line ATM interface using Python and Git in VS Code, implementing secure PIN validation and error handling.

Streamlined code logic for balance checks and transactions, supporting 100+ test cases with 98% pass rate.

Debugged Python scripts to optimize performance, reducing runtime errors by 20%.

Developed a Fake News Detection system using Logistic Regression and TF-IDF vectorization, achieving 95% accuracy with Scikit-learn and NLTK (Python).

Optimized NLP performance by 10% through hyperparameter tuning, text preprocessing, feature engineering, and model evaluation.

Implemented cross-validation techniques, improving model robustness on a 10,000+ record dataset.

Designed data pipelines for 5,000+ text entries, reducing processing time by 15%.

Collaborated in an Agile team of 4 to deploy models via REST APIs, enhancing real-world news classification.