About - Griffin Rush

About Me

I'm a Data Engineer specializing in scalable cloud data solutions and AI-powered analytics platforms. My work centers on building production data pipelines, intelligent agents, and automated workflows that turn raw data into decisions.

I'm skilled in Python, SQL, PySpark, Azure Databricks, Unity Catalog, MLflow, and LangGraph, with deep expertise in modern data engineering practices — from architecting Delta Lake pipelines to orchestrating multi-agent systems for governed, natural-language analytics.

🎓

Education

B.S. Informatics, Indiana University - Bloomington

Minor in Business & Financial Literacy

💼

Current Role

Data Engineer - L6, General Motors

July 2021 - Present

⚙️

Core Stack

Python, SQL, PySpark, Azure Databricks

Unity Catalog · MLflow · LangGraph

Career

Professional Experience

Data Engineer - L6

July 2021 - Present

General Motors

Develop and maintain production data pipelines using PySpark, SQL, and Python to aggregate and transform data stored in Databricks Unity Catalog, supporting business intelligence and advanced analytics initiatives across marketing and business units. Architect scalable data solutions on Azure Databricks leveraging Delta Lake, Unity Catalog, and Spark to process and curate large-scale transactional and event data from disparate first- and third-party sources. Collaborate with cross-functional analytics, data science, and business teams in a matrixed environment to translate requirements into robust data products that enable self-service analytics and ML modeling. Champion continuous learning and knowledge sharing within the data engineering team by documenting best practices, conducting technical sessions, and staying current with emerging technologies and industry trends.

PySpark SQL Python Azure Databricks Delta Lake Unity Catalog

Selected Work

Projects

Agentic Analytics Platform

Designed and deployed a multi-agent analytics platform integrating Databricks Genie, Unity Catalog, and custom LangGraph agents to enable governed natural-language querying of marketing datasets, campaign performance metrics, and experimentation results
Architected a custom LangGraph supervisor agent to orchestrate multiple specialized sub-agents, including vector search for semantic retrieval, externally-served agents (Genie) for structured experiment data, and in-code agents for dynamic code execution and programming assistance
Engineered robust prompt orchestration with system prompts and policy guardrails to ensure concise, grounded, and compliant responses while managing tool routing, SQL generation, and policy filters to map business questions to correct Delta tables and metrics
Deployed agent as an MLflow ResponsesAgent for seamless integration, version control, and end-to-end tracing of agent interactions and model performance

Automated Template Processing & Dynamic Content Generation

Automated end-to-end deck generation that reads governed PowerPoint templates from Volumes, populates text and image placeholders from source tables, and writes polished outputs, eliminating manual deck assembly
Preserved brand-consistent formatting by copying font attributes during text replacement, ensuring professional visuals without rework
Built a reusable parameterized workflow, including widgets for template, table, filters, and output, that lets teams generate tailored presentations without code changes
Improved reliability with exception-safe image handling and text/table fallbacks, keeping runs and paragraphs valid and reducing failed outputs