KIM DOOINN

I’m a Data Scientist with 3+ years of experience building data pipelines, analytics products, and AI engineering (LLMs), focused on improving data correctness and delivering decisions that teams can trust.

I’ve delivered scalable reporting systems and measurable business impact across product, supply chain, and prompt engineering—monitoring AI agents in production, designing evaluation metrics for quality and performance across LLM versions, and building systems that catch prompt or model regressions early.

In my work, I turn messy data into fast, interactive analytics: I migrated static Excel reporting into Power BI dashboards for real-time visibility, reduced reporting time by 80–100% with RPA/Power Query/SAP Script/VBA, and shipped a fine-tuned BERT “Product Feedback Analyzer” pipeline with end-to-end NLP automation and Docker deployment.

I also strengthen models throughout their lifecycle. I created a prompt optimization tool for an LLM migration, improving performance by 10–15% while cutting prompt optimization time by ~30%, and I support model development with hands-on ML practices—from MLflow-tracked experimentation and n8n workflows to Python/R analytics and survival analysis for churn insights.