Software Engineer with 8+ years of experience, specializing in Python-driven automation and machine learning. Experienced in building, evaluating, and optimizing ML models and data pipelines, with a strong track record of accelerating experimentation and improving workflow efficiency. Brings a growth mindset, strong curiosity, and a commitment to continuous learning to quickly adapt, expand capabilities, and drive ongoing improvement.
Developed and fine-tuned a transformer-based NLP classifier to detect food-related image captions using Hugging Face and PyTorch. Generated a synthetic labeled dataset using an LLM and leveraged transfer learning to accelerate model development. Deployed an interactive demo using Gradio on Hugging Face Spaces.
Explore Project →Developed and fine-tuned a multi-class image classification model using TensorFlow 2 and TensorFlow Hub to identify dog breeds from images. Implemented data preprocessing, visualization, and model evaluation with NumPy, Pandas, Matplotlib and leveraged transfer learning to accelerate training and improve classification performance.
Explore Project →Developed and optimized a supervised classification model using Scikit-Learn to predict heart disease risk from clinical patient data, achieving a 95% accuracy through model selection, cross-validation, and hyperparameter tuning.
Explore Project →Developed and optimized a regression pipeline using Scikit-Learn to predict bulldozer auction prices from structured equipment and sales data, leveraging data processing, feature engineering, and model optimization techniques. Evaluated model performance using Root Mean Squared Log Error (RMSLE) to ensure robust price predictions across varying price ranges.
Explore Project →Implementation of a basic word tokenizer.
Explore Project →