About Me
I'm a B.Tech. Data Science student at Mukesh Patel School of Technology Management & Engineering, NMIMS, with experience in data analytics and machine learning internships. I've worked on fraud detection, credential stuf…
I'm a B.Tech. Data Science student at Mukesh Patel School of Technology Management & Engineering, NMIMS, with experience in data analytics and machine learning internships. I've worked on fraud detection, credential stuffing detection, credit risk modeling, sentiment analytics, and AI-powered microservices using tools and technologies including Python, SQL, PyTorch, TensorFlow, LangChain, FastAPI, AWS, Azure, Spark, Power BI, Tableau, Git, MLflow, and Docker.
Experience
Business Analyst Intern
<p>Automated and deployed an ML-driven fraud detection framework using unsupervised anomaly detection, AWS Lambda, and S3.</p><p> Leveraged behavioral analytics, anomaly scoring, and alerting to uncover fraud patterns across large-scale authentication events.</p><p> Achieved 100% detection precision and less than 4.5% false positives on validated fraud cases.</p><p> Implemented a credential stuffing detection model using multi-temporal features, weighted velocity modeling, and ensemble machine learning techniques.</p><p> Achieved 90.54% precision and 97.27% recall against sophisticated automated attacks.</p>
Machine Learning Intern
<p>Built AI-powered microservices for a skills-first hiring platform using Azure OpenAI and FastAPI.</p><p> Developed a resume intelligence module.</p><p> Developed a labor market analytics pipeline.</p><p> Developed an NLP-based skills extraction engine.</p><p> Developed an adaptive assessment engine pairing LLM-generated questions with a custom ML-driven model calibration for personalized candidate evaluation.</p>
Data Analytics Intern
<p>Designed a credit risk modeling framework across multiple financial customer segments.</p><p> Applied statistical feature selection and model benchmarking to reduce feature space from 200+ variables to 10 key predictors.</p><p> Enhanced domain-specific credit risk models through feature engineering, ensemble methods, algorithm evaluation, and hyperparameter optimization.</p><p> Increased the Gini coefficient by 21%.</p><p> Improved predictive risk discrimination.</p>
Data Analytics Intern
<p>Improved sentiment analytics pipelines for social media data.</p><p> Leveraged text preprocessing, feature extraction, and Power BI visualizations to uncover engagement trends.</p><p> Enabled real-time insight generation and performance monitoring.</p>
PROJECTS
Credential stuffing
<p>Automated and deployed an ML-driven fraud detection framework using unsupervised anomaly detection, AWS Lambda, </p><p>and S3, leveraging behavioral analytics, anomaly scoring, and alerting to uncover fraud patterns across large-scale </p><p>authentication events, achieving 100% detection precision and