Looking for a job

Amulya Kumari

@amulyakumari

Senior Data Engineer at PWC building metadata-driven Spark and Scala pipelines for marketing data on GCP.

At PWC, I build metadata-driven Spark and Scala pipelines for marketing feeds, using Cloud Dataproc, Cloud Composer, and BigQuery. I moved ETL logic out of orchestration and into reusable transformations driven by metadata tables.

I’ve tuned Spark workloads, stabilized production jobs, and added validation, quarantine handling, and lineage tracking to help keep marketing data reliable. At Ford, I refactored marketing ETL across AWS and Azure, including Glue pipelines, Redshift models, and streaming integrations.

At Accenture, I worked on Spark and Scala marketing pipelines and also built healthcare data flows using AWS and Snowflake. I hold Azure, AWS, and Google Cloud data engineering certifications.

Experience

Work history, roles, and key accomplishments

Current

Senior Data Engineer

Current

PWC

Nov 2024 - Present (2 years)

Delivered a scoped Spark Scala marketing feed on Cloud Dataproc backed by Cloud Storage, wiring outputs into BigQuery under tight campaign launch timelines. Built a reusable metadata framework where Cloud Composer triggers Dataproc jobs that read column rules, filters, and join definitions from tables instead of hardcoded Scala maps.

Data Engineer

Ford

Mar 2022 - Nov 2024 (2 years 8 months)

Delivered a time-boxed refactor of legacy marketing ETL into AWS Glue jobs, separating orchestration from logic and stabilizing daily runs under tight reporting SLAs. Reworked core data warehouse structures in Amazon Redshift using star schemas and proper keys, aligning with marketing analytics requirements and avoiding hidden ETL logic in scheduling layers.

Big Data Engineer

Accenture

Aug 2019 - Mar 2022 (2 years 7 months)

Delivered a scoped Spark and Scala batch pipeline for marketing feeds, wiring into existing Azure storage and warehouse layers without changing downstream contracts. Reworked legacy ETL flows so Azure Data Factory only scheduled jobs, moving all business rules into Spark transformations and shared Scala libraries for clearer ownership.