Narikinabiilli Jyotsna

Results-driven Data Engineer with 10+ years of experience designing, building, and scaling large-scale data pipelines for banking and financial services organizations, based in Hyderabad. Proven expertise in Python, SQL, PySpark, and Spark SQL, with a track record of re-platforming legacy mainframe workloads onto distributed Spark pipelines to cut end-to-end runtime by over 95%. Skilled across the modern data stack — Hadoop, Hive, Cassandra, and Microsoft Azure — with hands-on experience in Git-based version control and Jenkins CI/CD for reliable, production-grade pipeline delivery.

Known for improving data integrity, query performance, and pipeline reliability while partnering closely with cross-functional business and analytics stakeholders.

Software Engineer II B - Bank of America - Hyderabad, India

(2020-04)

• Re-engineered legacy mainframe Gap Account generation and data sanitization workflows into distributed PySpark pipelines, reducing end-to-end runtime by over 95% and fully automating previously manual processes.
• Built orphan-record detection logic within account data pipelines to systematically surface data inconsistencies, strengthening downstream data integrity and reducing manual reconciliation effort.
• Partnered with cross-functional teams to define data readiness requirements for low-lying environments supporting the account infusion process, translating business needs into pipeline logic.
• Delivered measurable infrastructure cost reductions by decommissioning legacy mainframe workloads in favor of scalable Spark-based processing.
• Used Git-based version control and Jenkins CI/CD pipelines to manage code releases and deployments for PySpark data pipelines, improving release consistency and reducing deployment errors.
• Building an NLP-driven Cross-Domain Schema Intelligent Platform (knowledge base) using spaCy to automatically detect and match schemas across different data domains, combining data preparation, feature engineering, and model application; achieving ~80% suggestion accuracy against business-defined matching requirements to date.
• Designing and building a Synthetic Data Service with FastAPI endpoints (3 to date), exposed via MCP (Model Context Protocol) so clients can request data directly through GitHub Copilot — cutting client data-request turnaround from an approximate 4-week manual process to near-real-time self-service.

Associate Projects - Cognizant — Client: JP Morgan Chase - Hyderabad, India

(2017-11 - 2020-03)

• Designed and built scalable data pipelines on Hive (Text & Parquet formats) to process large enterprise datasets, identify unused attributes, and optimize storage, reducing cold-data footprint by 25–30%.
• Developed Spark and Hive SQL jobs to detect and eliminate duplicate tables, columns, and records, improving downstream data quality by nearly 35%.
• Built ingestion pipelines feeding processed results into Cassandra and Hive, enabling near-real-time analytics for multiple business teams.

Senior Software Engineer - Capgemini — Clients: Element Financial, GE Transport, GE Aviation - Bangalore, India

(2014-01 - 2017-10)

• Migrated Sybase reporting workloads to Cassandra and Spark, improving BI query response time from ~3 minutes to ~6 seconds and enabling near-real-time analytics.
• Engineered optimized Cassandra data models using RDBMS profiling, grain consolidation, and performance tuning to support high-volume analytical queries.
• Built Spark SQL ingestion pipelines from FileoDB into BI reporting layers while tuning Cassandra clusters for low-latency, scalable reporting.
• Developed end-to-end BI pipelines moving raw Excel data into an Oracle data warehouse, including data cleansing, normalization, and validation.
• Led data analysis, technical documentation, and client collaboration to translate business requirements into scalable BI/data solutions.

Bachelor of Engineering - Electronics and Communication Engineering - Jawaharlal Nehru Technological University, Kakinada