At Accenture, I architected an agentic data engineering platform with Python, LangGraph, Databricks, and Snowflake. Its LLM-powered SQL and PySpark workflows validate and self-correct execution, reaching 93% successful query execution.
I also developed agents for schema discovery, data profiling, and pipeline diagnosis, with human approval and policy guardrails for safer changes. I optimized Spark workloads, reducing pipeline runtime by 33% and compute cost by 24%.
At Shopify, I built a causal ML experimentation platform using Python, Spark, and EconML to measure incremental outcomes. My uplift models increased incremental conversion by 18%, and I processed 30M+ daily events for large-scale causal modeling.
