Aaron Clewis

I've built practical AI estimation and workflow automation for Collision Champs, replacing fragmented estimate and CRM processes with a custom operating stack connecting GoHighLevel, AWS-based CCC syncing, customer communication, and sales reporting. I evaluated and integrated Tractable AI, built Python photo-quality checks and guided website prompts, and designed fallback paths so poor photos, missing VINs, and AI failures would not lose leads. Previously, I built and documented a BASEL loss forecasting model at Wells Fargo with an estimated $1B benefit over the existing EAD model over five years. At CoreLogic, I developed production alternative-lending credit models using logistic regression, Python, SAS, Teletrack attributes, and derived features. My background also includes demand forecasting and reporting automation at JCPenney, marketing analytics at Citi and Javelin Direct, and economics instruction at the University of North Texas. I work across Python, SQL, SAS, VBA, cloud platforms, statistical modeling, forecasting, and operational automation.