Work Experience

Key Contributions:

• Model Development in Python: Developed 23 regression models in Python, each optimized for its assigned task. Models included advanced regressors such as Stacking Regressors (leveraging ensemble methods), Gradient Boosting, Ridge Regression, Random Forest, and K-Nearest Neighbors. Each model was evaluated using train-test splits and cross-validation to determine the optimal combination of regressors and features.
• Synthetic Data Integration: Generated and incorporated synthetic data to improve the generalization and robustness of the models, particularly for underrepresented scenarios.
• Feature Engineering: Engineered domain-specific features to improve cost comparability across diverse project configurations. This resulted in enhanced model performance, including improved R², MAE, and MSE scores.
• Application Integration: Serialized 23 task-specific models into .pkl files, integrating them into a unified application with a Python backend and an HTML-based user interface. Each model uses its optimal features to generate precise predictions, while the centralized application provides seamless input handling and accurate cost estimation.
• Deployment via Azure DevOps: Deployed the finalized application and models to a live production environment through Azure DevOps.

This visualization demonstrates the performance of a specific task's model, achieving an R² score of 0.86 and outperforming historical estimates 85.7% of the time. It compares the predicted costs (model output) with historical estimates and actual project costs:

Key Contributions:

• Workflow Automation: Built a Power Automate flow to manage user submissions, dynamically handle photo uploads, generate unique post IDs, and store images in a shared internal folder.
• Data Integration: Connected Microsoft Forms, Excel storage, and Viva Engage to enable structured data handling and automated community email notifications.
• Flow Complexity: Utilized JSON parsing, loops, and conditional logic to process files, append them to arrays, and compose messages for Viva Engage posts.

Key Contributions:

• Visualization: Created R-based visualizations to improve the interpretation and communication of mass balance metrics.
• Advanced Analytics: Analyzed mass balance data using Minitab and Tableau, applying statistical metrics such as mean squared error and R² to assess trends and identify improvement opportunities..