Dhruv Pai

• Developed and trained a LightGBM based machine learning model on a company-provided dataset containing 1M+ feature relationships.
• Built the end-to-end modeling pipeline including feature engineering, training, evaluation, and validation, achieving ~94% accuracy with strong precision and recall.
• Collaborated with team members to refine feature insights and model design; presented model approach and results to an evaluation jury.

• Collaborated in a cross-functional team to build an AI powered chat system for customer interaction.
• Designed and implemented NLP pipelines for sentiment analysis and text classification on customer reviews and feedback.
• Utilized NLTK and spaCy for text preprocessing, feature extraction and model development.

• Built an ML-based intrusion detection system for encrypted network traffic using flow-level and temporal features, without inspecting packet payloads.
• Trained classification models (Lighgbm,xgboosted)on 1M+ network traffic records with 30+ engineered features, achieving strong precision, recall, and ROC-AUC.
• Applied SHAP to interpret feature importance and explain model decisions and presented system design and results to an academic jury.