Joshua McCourt

At Ideagen, I resolved a production AWS Batch outage affecting 11 consecutive nightly runs and shipped a validated Terraform fix with zero production changes. I also built and deployed a stateless AWS connector service so an LLM agent could analyse alarm history and post recurring-alert roundups to Microsoft Teams.

I've built end-to-end machine-learning systems spanning NLP fraud detection, electricity price-spike prediction, and production-ready model monitoring. My electricity-price pipeline achieved ROC-AUC 0.9966 and included time-aware validation, probability calibration, schema validation, drift monitoring, versioned artefacts, and shadow-deployment comparisons.

I turn multi-source data into decision-ready insight using Python, SQL, AWS, Terraform, and Tableau. For an Instacart analytics project, I built a SQL ETL pipeline processing 30M+ transactions, designed a star schema, and delivered cohort, retention, product-affinity, and customer-behaviour analysis.

My background in Economics, Data Science, Mathematics, and Economics combines statistical reasoning with practical engineering. I'm particularly interested in production ML systems, applied AI agents, distributed systems, and scalable cloud data platforms.