RAMYA BALIGA-BHATT

Data Engineer | AI & Agentic Systems
● Open to Work US Green Card · No Sponsorship Required
San Francisco, CA +1 415 769 9004 ramyabaligabhatt@proton.me linkedin.com/in/ramyabaliga

Professional Summary

Data Engineer with 7+ years building production-grade ETL/ELT pipelines and cloud data warehouses at enterprise scale. Specializes in AI-driven data systems with hands-on experience in autonomous error detection, LLM-powered interfaces, and full-stack delivery from ingestion to dashboards. Proven track record of reducing pipeline runtime by 40% and eliminating 10+ recurring failure points across 50+ databases.

Technical Skills

Autonomous Agent Systems

Multi-Agent Orchestration, Autonomous Error Detection, Self-Healing Pipelines, Agent Decision Systems, Failure Recovery, Tool Use & Function Calling

Production LLM Engineering

Retrieval-Augmented Generation (RAG), LLM Evaluation Frameworks, Multi-Model Routing & Fallback, Prompt Engineering, Text-to-SQL, Context Window Management

AI Reliability & Observability

Production AI Monitoring, Pipeline Health Dashboards, Automated Quality Gates, Data Lineage Tracking, Hallucination Mitigation, Latency Optimization

AI Infrastructure

Vector Databases & Semantic Search, Containerized AI Deployment, API Design for LLM Systems, Cloud Data Platforms (Azure, AWS, GCP), DAG Orchestration (Dagster, Airflow)

Data Engineering Core

Python, SQL (MSSQL, Spark SQL, PL/SQL), PySpark, ETL/ELT Pipeline Design, Cloud Data Warehouses (Azure Synapse, Snowflake, Databricks), Data Modeling (Star Schema), DuckDB, Parquet

DevOps & Delivery

Docker & Container Registry, CI/CD Pipelines (Jenkins), Kubernetes, Git Version Control, Agile/Scrum (CSM), Infrastructure as Code, Zero-Downtime Deployments

Featured Projects

AskMyData — Text-to-SQL Data Exploration App GitHub
  • Built LLM-powered text-to-SQL application for querying World Bank datasets in natural language, supporting multi-provider failover (Claude, GPT-4o) for 99.9% uptime
  • Engineered Parquet + in-memory DuckDB pipeline achieving sub-second analytical queries on 100M+ records through optimized columnar storage
  • Implemented Plotly auto-charting and searchable dataset catalog; deployed on Streamlit with session-state management
Self-Healing Data Pipeline GitHub
  • Built autonomous ETL error-recovery framework combining regex rules engine (80%+ auto-fix rate) with LLM fallback (GPT-4o-mini / Claude Haiku) for novel failures
  • Implemented 8 severity-gated remediation strategies (schema drift, credential refresh, backoff, null quarantine) with automated dispatch
  • Developed self-learning SQLite knowledge base that auto-promotes recurring errors to permanent rules; integrated with Dagster via failure hooks
Developer Portfolio Website GitHub
  • Built responsive, animated portfolio in Next.js 16, React 19, and Tailwind CSS 4 with config-driven, typed content architecture
  • Implemented dark/light/system theming, shadcn/ui components, scroll-triggered animations, and full mobile responsiveness

Professional Experience

Data Engineer — Johnson & Johnson
Jun 2022 – Dec 2024
Limerick, Ireland | Enterprise Data Platform Team
  • Led backend redesign for legacy platform — re-architected database structure and cleansing rules across most tables
  • Provisioned and managed Azure infrastructure end-to-end — resource groups, VMs, Data Factory (ADF) pipelines, and App Service Environments — supporting production ETL workloads
  • Built and maintained Databricks notebooks in Python for data processing, managing cluster access, secrets, and job scheduling across QA and production environments
  • Wrote and optimized complex SQL queries (3,000+ lines) for critical reporting and data transformation logic
  • Built and owned 15+ ETL pipelines moving 500GB/day from 20+ source systems into Azure data warehouses and lakes, cutting manual data-prep time by 35%
  • Re-architected pipeline performance across 50+ databases, data lakes, and data warehouses, reducing average run time by 40% and eliminating 10 recurring failure points
  • Designed data models and schemas supporting 50+ downstream reports, cutting average query time by 30%
  • Rolled out data quality checks, versioning, and lineage tracking across 15+ pipelines, cutting downstream data errors by 45%
  • Led a solo, end-to-end Azure subscription migration — cloning ADF pipelines, Databricks environments, and production/QA infrastructure into a new subscription, and cutting over on go-live day with minimal disruption
  • Deployed containerized applications via Docker and managed the container registry; administered versioning and configuration across Azure services
  • Implemented data security controls and pipeline monitoring, identifying and resolving 20+ performance bottlenecks before they impacted production
  • Maintained version-controlled documentation for 15+ pipelines and schemas in Git, cutting new-hire ramp-up time by 25%
Data Analytics Developer — Johnson & Johnson
Jun 2021 – Jun 2022
Limerick, Ireland | Supply Chain Analytics
  • Built ingestion and preprocessing workflows across 10+ databases, APIs, and log sources, feeding 30+ downstream analyses per week
  • Developed 20+ Tableau and Power BI dashboards used by 50+ stakeholders to track real-time supply chain and quality metrics
  • Conducted statistical testing and exploratory data analysis on 15+ datasets, surfacing 10+ actionable insights adopted by business leadership
  • Proposed data-driven recommendations on 10+ business problems, with 60% adopted into production workflows
Business Data Analyst — Johnson & Johnson
Jan 2021 – Jun 2021
Limerick, Ireland | Business Intelligence
  • Translated complex data findings into actionable insights for 20+ non-technical stakeholders across 5 business teams
  • Optimized ETL processes across databases, CRM systems, and survey data, reducing processing time by 25%
  • Managed 8+ data projects end-to-end, meeting 100% of stakeholder deadlines
  • Built 50+ dashboards in Excel, Tableau, and Power BI, adopted as team's standard reporting format
Software Developer — Johnson & Johnson
Sep 2019 – Jan 2021
Limerick, Ireland | Mobile & Web Development
  • Built complete front end of iOS app ("Beerse One Way") delivering guided tours to 5,000+ daily site visitors
  • Developed full-stack features using Node.js and React.js, contributing to internal application serving 200+ daily users
Junior Software Developer — Awnics
Mar 2016 – Jul 2017
India
  • Built front-end and back-end features for in-house Node.js/React.js application

Education

M.Eng, Computer Engineering — 1:1 Honors
University of Limerick
2017 – 2019
Continuous Professional Development, Engineering — 1:1 Honors
Technical University Dublin
2019 – 2020

Certifications

Microsoft Certified: Azure Fundamentals Microsoft Certified: Azure Data Fundamentals Scrum Alliance Certified ScrumMaster (CSM)
LinkedIn Learning Certificates
Generative AI for Business Leaders Introduction to Prompt Engineering for Generative AI Prompt Engineering: How to Talk to the AIs Cloud Concepts: Determining Your Cloud Strategy Advanced SQL for Query Tuning and Performance Optimization DevOps Foundations Learning Power BI Desktop Cloud Architecture: Core Concepts Cloud DevOps Concepts: Understanding Processes and Services Analyzing Big Data with Hive