"
Hello!
I'm VineethSanke
I'm a GenAI/ML Engineer with 5+ years of experience building production-grade data infrastructure that powers real ML systems. I hold a Master of Science in Computer Systems Analysis from Pace University, and have built my career specializing in end-to-end MLOps on AWS from high-throughput feature ingestion pipelines using SageMaker and AWS Glue, to RAG framework deployment and LLM guardrail systems.
I’m My background spans financial and healthcare data at scale, with hands-on experience reducing model inference latency by 50%, eliminating data drift, and compressing data-to-model lifecycle times by 40%. I've worked across Spring Health, Discover Financial Services, and Deloitte delivering clean data in, reliable models out, monitored in production.
Pace University
New York, NY
GPA: 3.92
Coursework: Data Analytics, Advanced Machine Learning, Fundamentals of Information Security, Software Design Quality, Web Application Development, Information Visualization, Usability Engineering and Social Media Analytics.
Vellore Institute of Technology (Business School)
Vellore, India
CGPA: 8.25/10
Scalable LLM & RAG Infrastructure: Architected high-throughput feature ingestion pipelines using AWS Glue and Amazon SageMaker, processing 5M+ records into optimized Parquet formats to power a RAG (Retrieval-Augmented Generation) framework, reducing model inference latency by 50%.
NLP Text Preprocessing & Feature Engineering: Built scalable text preprocessing pipelines using AWS Glue and SageMaker, transforming 5M+ raw clinical records into tokenized, embedding-ready Parquet formats. Reduced data preparation overhead by 40%, accelerating downstream LLM training and inference workflows.
Enterprise MLOps & Orchestration: Developed automated CI/CD/CE pipelines using SageMaker Pipelines and AWS Step Functions, ensuring 100% reproducibility of LLM prompts, training datasets, and feature sets across development and production environments.
Credit Risk & Fraud Analytics: Built SQL and PySpark anomaly detection models to flag suspicious transaction patterns across 10M+ monthly credit card records, reducing false positive rates by 22% and supporting near-real-time fraud intervention workflows.
Complex Financial ETL: Designed AWS Glue and PySpark pipelines to ingest and standardize 3M+ records from disparate source systems, improving data readiness for actuarial risk modeling by 30%.
Business Intelligence & Reporting: Developed Tableau and Power BI dashboards tracking delinquency rates, charge-off trends, and approval rate KPIs giving risk and product leadership a single source of truth for lending strategy decisions.
Snowflake Data Governance: Implemented zero-copy cloning to enable secure cross-functional data access across risk, compliance, and product teams, reducing storage overhead by 60% while maintaining strict SOX and Basel III audit controls
Stakeholder Data Translation: Partnered with risk, compliance, and product teams to convert business requirements into analytical frameworks, delivering ad-hoc and recurring reports that directly informed credit policy adjustments.
