Ash xpxp

I’m an AI & Data Engineer focused on production-grade data reliability and measurable performance improvements. At Deloitte (USI), I migrated a massive legacy PostgreSQL system into a modern Databricks/Delta Lake architecture (Bronze/Silver layers) and engineered PySpark pipelines to transform deeply nested XML into clean, query-ready tables.

I also build audit-ready data models and automation: I implemented optimized MERGE INTO upserts with partition pruning and strict SCD Type 2 tracking for financial accuracy, and migrated 100+ complex tables with 5M+ records. I led an end-to-end automated testing and data-validation pipeline that reduced QA time from ~2 hours to ~30 seconds per run, eliminating human error.

Earlier, I delivered ETL/ELT infrastructure and fault-tolerant workflows for the Utah Department of Transportation using BigQuery, AWS S3, SSIS, and SSMS—validating data, quarantining error rows, and applying efficient SCD Type 1 strategies. As an LLM Tech Intern, I designed a hierarchical multi-agent LangChain architecture and built an Angular frontend to render real-time agent execution traces, alongside hands-on ML/NLP projects in healthcare and fraud detection.