Hi there, I’m Saurabh.

Here’s my profile description:

• Skilled Data Scientist with 5+ years of experience in Data Extraction, Data Modelling, Data Wrangling, Statistical Modeling, Data Mining, Natural Language Processing, Machine Learning, and Data Visualization. • Track record of driving business growth through data-driven strategies. • Expertise in A/B testing, recommendation systems, and user engagement, resulting in a 15% increase in footwear trades and 20% higher user engagement. • Proficient in Python, AWS, SQL, and Docker, with a background in data analysis, statistical modeling, and manipulation using R. • Reduced operational costs by 30% and improved system reliability by 40% through AWS infrastructure design and deployment. • Experienced in ETL processes, Kubernetes, CI/CD pipelines, and container orchestration for efficient data management and deployment. • Recognized for leading impactful credit card marketing campaigns, achieving a 16% increase in open rates, and contributing to a $3 million reduction in fraud losses in the first year for a prominent US banking client.

Experience

Enstructure : Data Scientist - Full Stack (Nov’23-*)

• Engineered an end-to-end data pipeline for a Real Estate client leveraging AWS Glue, incorporating custom Lambda triggers
• Elevated decision-making with automated dashboards on AWS Quicksight, providing dynamic insights into critical metrics

Trinet : Product Data Scientist (Jun’23-Oct’23)

• Led cross-functional collaboration to optimize RISK-NAICS code assignment, integrate classification models, reduce manual processing time by 25%, and achieve a 30% increase in client acquisition through refined risk evaluation accuracy for dynamic pricing of healthcare insurance
• Leveraged BERT-based NLP embeddings and topic modeling to analyze and classify customer support tickets, identifying critical factors driving customer attrition
• Integrated Salesforce with external systems using web services and APIs, improving data accuracy by 40% and reducing data entry errors by 50%

Tradeblock : Graduate Capstone Project - Data Science (Jan’23-May’23)

• Achieved a notable 15% increase in footwear trades and 20% higher user engagement through Python-based A/B testing and custom recommendation systems, all while maintaining 98% of the original variance with PCA.
• Managed Docker deployment, integrated AWS services with on-premise resources for an 8% cost reduction, optimized ETL processes, and skillfully utilized Kubernetes for container orchestration.
• Skills: Pattern Recognition · Statistical Modeling · A/B Testing · Amazon Web Services (AWS) · Data Science

EXL Service : Data Scientist (Aug’19-Apr’22)

• Led a 6-member model development team to predict Gross Credit Loss for a fortune-100 US bank with 98% accuracy
• Deployed credit card cross selling model using K-Means customer segmentation, enabling the client run targeted promotions
• Created a credit card fraud detection model, using Logistic Regression, for a leading US banking client, helped in reducing fraud losses by 3 million USD during the first year of its deployment
• Implemented a Decision Tree algorithm with hyperparameter tuning, helped to reduce customer churn rate by over 40%
• Performed regression on the customer data to evaluate credit refinement, leading to a decrease in Net Credit Loss by over 6%
• Designed and optimized Data Warehousing Architecture for a US based bank which led to a reduction in query time by 15%

Projects

Please click on the Headline to be redirected to my Portfolio Page. You can also click here