Mrunmayee Kulkarni

About Mrunmayee Kulkarni

Build high-performance scalable data warehouses. Design and launch efficient & reliable data pipelines to move and transform data Interact with product owners and end users to understand their needs and solve issues. Deliver data with the highest quality through rigorous checks. Have excellent data modeling skills to understand the nuances of various dimensions and metric types in the warehouse. Design workflows to ingest, load, and present new data sets for users. Provide active support, and be on rotation for on-call support on production pipelines Define and manage SLA for all data sets in allocated areas of ownership. Design data models for optimal storage and retrieval. Work with the production engineering/infrastructure team to drive resolution to production issues.Skills: AWS, GCP Data pipelines Oozie, airflow orchestration Programming Languages: Python, Java, C, C++, Pig, SQL, Shell Databases: BigQuery, Hive, Microsoft SQL Server, MongoDB, Spanner, Apache Druid, MySQL, PostgreSQL Frameworks: Hadoop, Oozie, Airflow, Python Flask, PySpark, Spark, Kafka, MapReduce Cloud: AWS, GCP, Redshift, Glue, S3, Athena, Cloud Watch, S3 Buckets Tools/Services: Google Dataproc, Composer, Looker, Jira, Git, Hue, BI, Selenium, JUnit, Docker, Kubernetes, CI/CD Web Technologies: REST API, PHP, HTML, CSS, React, JavaScript

Experience

1. Data Engineer - Genai Customer Service

   Amazon

   Jan 2025 — Present · US

   Designed and built large-scale data pipelines processing TB-scale customer interaction data daily, powering Amazon\'s AI-driven CS products including Rufus LLM integration with chatbot, Help Pages, and CS Homepage. Architected ETL workflows on AWS (Glue, S3, Redshift) for LLM training dataset curation and RAG-based content retrieval systems. Built customer journey tracking pipelines capturing resolution metrics and help page effectiveness signals. Developed QuickSight dashboards providing visibility into customer behavior patterns, self-service resolution rates, and LLM performance metrics, enabling content teams to improve help page effectiveness by 60% and reducing average resolution time. Implemented data frameworks with feedback loops to identify LLM improvement opportunities, contributing to continuous model optimization across millions of daily customer interactions.

Education

• K.K.Wagh institute of Engineering Education & Research

  Bachelor of Engineering - BE, Information Technology

• Binghamton University

  Master's degree, Computer Science