Headline

ML | NLP | Data Science | PyTorch | HuggingFace | Ensemble Techniques | Scikit-learn | Optuna | IIT Madras | Open to Intern & Full-Time

About

ML and Data Science engineer with hands-on experience building end-to-end intelligent systems — from exploratory data analysis and feature engineering to deep learning architectures and model optimization on real-world datasets. Most recently fine-tuned RoBERTa for fake job detection, combining text embeddings, numerical features, and categorical embeddings into a multi-modal classification system — with Focal Loss, layer-wise learning rates, gradient clipping, and threshold optimization. Other projects include credit card fraud detection comparing Logistic Regression, Random Forest, and XGBoost (PR-AUC 0.77, Recall 0.82 across 284K transactions) and asteroid hazard classification using Optuna Bayesian search on NASA/JPL data. Comfortable across the full ML workflow — data cleaning, EDA, feature engineering, imbalanced data handling, model development, hyperparameter tuning, and performance evaluation using Precision, Recall, F1, ROC-AUC, and PR-AUC. Stack: Python · PyTorch · HuggingFace · Scikit-learn · XGBoost · Optuna · Pandas · NumPy · SQL IIT Madras · Open to ML / Data Science / AI internships and junior roles · India / Remote

Location

Delhi

AI stack

cursor

opencode

Work experience

ML Analyst at Self-Employed, Mar 2026 - Present · 6 mos, Building end-to-end machine learning solutions across domains including NLP, fraud detection, and classification. Key work includes: -Fine-tuned RoBERTa-base transformer for fake job detection using multi-modal fusion combining text embeddings, numerical, and categorical features -Built deep learning fraud detection system on highly imbalanced financial transaction data using Focal Loss and class-aware optimization -Developed asteroid hazard classifier using NASA/JPL data with Bayesian hyperparameter search via Optuna Proficient in PyTorch, HuggingFace Transformers, Scikit-learn, Optuna, Pandas, and NumPy

Education

12 pass, Information Technology, Central Board of Secondary Education

programming and data science, Information Technology, Indian Institute of Technology, Madras