This is my profileContact via CazVidInderjeet
Data Scientist
Contact
Competencies
Demonstrated experience with Python.
Demonstrated experience with C++.
Demonstrated experience with SQL.
Demonstrated experience with Java.
Demonstrated experience with R.
Education
Summary
Data Scientist open to new opportunities
Experience
Research Experience — Machine Unlearning and Bayesian Learning-to-Forget
Recovered Bayesian linear-regression posteriors after deleting 50% of training data for machine unlearning. Compared exact and approximate Bayesian updates to evaluate posterior recovery under targeted data deletion.. Benchmarked machine-unlearning methods on CelebA-HQ with Grad-CAM to measure utility retention and forgetting.. Evaluated fine-tuning, noise injection, and layer reinitialization as fast-unlearning strategies on CelebA-HQ.
Data Scientist, Data Science Team
Built predictive ML models over millions of high-frequency sequential clickstream events to forecast next meaningful user actions from noisy event sequences and support hierarchical process discovery for a Fortune 50 client.. Developed a large-scale process-intelligence pipeline to transform raw clickstream events into user tasks, subprocesses, and hierarchical process maps across multiple abstraction levels for downstream modeling and analysis.. Designed algorithms to identify key business attributes and group semantically equivalent attributes from raw clickstream data, reducing client-side manual configuration by >80% across business processes and deployments.. Applied VLMs, OCR, and community detection to enrich event data and improve process-map interpretability.
Software Engineering Intern
Engineered Spring Boot Java pipelines to transfer data from DynamoDB into the central data lake for fast data access.. Integrated SNS and SQS to synchronize live user data with SQL tables powering internal analytics charts.