AKANKSHA RAJVANSHI

Data Science Workbench (Internal Product)

• Developed an end-to-end AutoML pipeline to accelerate ML workflows.
• Engineered feature selection, anomaly detection, data transformation, and dimensionality reduction modules.
• Built classification and regression models using Optuna and Scikit-learn (Random Forest, Stacking, Naive Bayes, etc.).
• Integrated Confusion Matrix, AUC-ROC, and RMSE visualizations for model evaluation.

Data Reconciliation – NCM Associates

• Developed a Python-based high-performance reconciliation tool using hashing and Jaro-Winkler matching.
• Reduced comparison time significantly while generating detailed mismatch reports.

Risk and Finance (Citizens Bank)

• Assisted in migrating data from IBM DataStage to Talend, minimizing data loss and improving process efficiency.
• Conducted an in-depth analysis of ETL processes, documenting job flows, data transformations, and source systems.
• Collaborated with cross-functional teams to ensure a seamless transition, reducing migration errors.
• Analyzed source-to-target mappings, identified migration challenges, and supported testing to ensure data accuracy and consistency.