Data Scientist with 5+ years of experience delivering Machine Learning and NLP solutions
Hyderabad, Telangana, India
#OpenToWorkData Scientist with 5+ years of experience delivering Machine Learning and NLP solutions across financial, healthcare, and telecom domains. Built predictive models achieving up to 85% accuracy, automated large-scale reconciliation workflows, reducing manual effort by 60%, and enhanced information retrieval precision by 30% using advanced NLP techniques.
I am looking for a Data Scientist /AIML Engineer role where I can work on Machine Learning, NLP, and data-driven solutions. I am interested in opportunities that allow me to apply my experience in Python, Predictive modelling, and automation while also enhancing my technical and problem-solving skills.
Finite HR & IT Solutions
• Collaborated with cross-functional stakeholders to translate business requirements into scalable ML solutions. • Streamlined DQI reconciliation, reducing manual validation effort by 60%. • Processed 1M+ IP records monthly using Python-based automation pipelines.
Cognizant
• Designed ML-driven automation improving operational efficiency by 35%. • Built a secure PDF redaction system ensuring 100% document access compliance. 1. Lotus Patient Safety – PDF Access Restriction • Systematized secure redaction of 5K+ medical documents with zero data leakage. • We are providing the restrictions to the documents, such as editing, printing, etc. • Implemented OpenCV-based image processing techniques for detecting sensitive information regions in scanned documents before redaction. • Developed deep learning workflows using TensorFlow/PyTorch to improve document classification and processing accuracy.
Wipro
• Developed classification models achieving up to 85% accuracy. • Improved model performance by 20% through EDA and feature engineering. Loan Return Prediction • Evaluated Logistic Regression, Decision Tree, and Random Forest models using cross-validation and ROC-AUC, selecting the best-performing model (0.86 ROC-AUC) for deployment. • Reduced dimensionality by 40% using PCA while maintaining performance. • Optimized model performance using GridSearchCV and cross-validation techniques to improve generalization. • Deployed the trained model using REST API for internal business users, enabling real-time loan risk prediction. • Enabled data-driven loan risk assessment supporting improved lending decisions.
CONIN Technologies
• Developed classification models achieving up to 85% accuracy. • Improved model performance by 20% through EDA and feature engineering.
M. Tech · Computer Science & Engineering
B. Tech · Computer Science & Engineering
