Open to opportunities

Sunandha Vipin Dev Kumar

@sunandhavipindevkuma

Message

Data science and cloud engineer focused on building scalable, LLM-powered analytics pipelines.

United States

Message

What I'm looking for

I’m looking for an entry-level data science/ML role where I can build reliable pipelines and dashboards, and apply LLM/RAG systems with evaluation rigor—shipping cloud-native solutions with clear governance and measurable impact.

I’m a data science professional who’s built ingestion, analytics, and ML systems end-to-end—using Python, SQL, and Tableau to turn messy records into fast, trustworthy decisions. In my internship at Next Chapter Property Solutions, I engineered Python/OCR/Pandas pipelines that cut manual processing by 90% and designed Tableau dashboards that reduced insight turnaround from weeks to minutes.

I also work deeply with AI/ML and production-minded architecture: I defined data quality rules, generated synthetic data for testing, and architected a GCP cloud-native ML pipeline with governance and deployment blueprints. Through projects like Contract Risk Auditing and my LLM-powered clinical risk assistant, I combine RAG, vector search, knowledge graphs, and rigorous evaluation to deliver measurable performance gains.

Experience

Work history, roles, and key accomplishments

Data Science & Systems Intern

Next Chapter Property Solutions

May 2025 - Jan 2026 (8 months)

Engineered Python, OCR, and Pandas ingestion pipelines for tax-sale records, reducing manual processing by 90%. Built SQL data models and Tableau dashboards, cutting insight turnaround from weeks to minutes and improving acquisitions workflow efficiency by 40%, while implementing automated anomaly detection with 100% stakeholder-aligned reporting coverage.

Python Pandas SQL Tableau Data Modeling Anomaly Detection Data Pipelines Data Visualization Segmentation

Data Science & Cloud Engineer Intern

Hexaware Technologies

Jun 2025 - Jul 2025 (1 month)

Defined 17 data quality rules and generated 500+ synthetic cheque images using Python, PIL, and Faker, increasing ML testing coverage by 3x. Built an end-to-end GCP cloud-native ML pipeline with governance and deployment blueprints, improving synthetic data fidelity via fraud injection and documentation that reduced cross-team knowledge transfer time by 40%.

Python Faker Synthetic Data Generation Data Quality GCP Cloud Native Governance Fraud Injection