Harika Sunkara

Developed batch ETL pipelines on Azure Databricks using PySpark and Azure Data Factory to integrate trading, customer portfolio, and market data into analytics-ready datasets. Built Delta Lake Bronze/Silver/Gold transformations and implemented reconciliation and schema validation to improve reporting and ML data reliability. Designed fact and dimension tables in Azure Synapse to support portfolio performance and risk dashboards, and tuned Spark jobs for performance.

Built batch ETL pipelines with Azure Data Factory, Azure Databricks, PySpark, and AWS Glue to ingest healthcare claims, enrollment, provider, and clinical data from databases, REST APIs, and flat files. Implemented Medallion-style Bronze/Silver/Gold layers in Delta Lake with schema validation, deduplication, and business transformations for analytics and predictive modeling. Created incremental ingestion and CDC workflows for eligibility, claim status, and provider changes, with data quality checks and monitoring aligned to PHI handling.

Wrote Python and SQL scripts to clean, transform, and validate structured datasets under guidance from senior engineers. Supported ETL data checks including null handling, schema checks, and duplicate identification, and produced SQL reconciliation outputs for team review. Documented pipeline steps, data issues, and test results to support reliable downstream processing.

Built a Python translation app using the OpenAI GPT-4o API with a Streamlit UI for bidirectional English–Spanish translation. Designed prompt templates to preserve context, tone, and idiomatic phrasing. Integrated REST calls with JSON handling and implemented error handling, then iterated on prompts using sample sentence evaluations.