Arseniy Korsun

Senior Data Scientist @Align Technology

Darmstadt, DE

Focus areas: Natural Language Processing (NLP), Large Language Models (LLMs), applied ML for healthcare and business problems.Key responsibilities & achievements: • Multilingual sentiment & dissatisfaction detection • Designed and deployed a multilingual NLP model to analyze text feedback from doctors. • Handled data in multiple languages (English, German, Spanish, Japanese, etc.). • Improved accuracy of dissatisfaction detection by >20% compared to rule-based baselines. • LLM-based synthetic data generation • Before the release of ChatGPT, built a proprietary LLM pipeline to generate medical text data for rare/risky treatment cases. • Enhanced model robustness and coverage in low-data scenarios. • Resulted in significant reduction of manual annotation costs. • Scalable ML on big data • Worked extensively with Databricks and Apache Spark for distributed data preprocessing, model training and evaluation. • Developed pipelines for handling millions of patient records and clinical notes while ensuring compliance with privacy regulations.Cross-functional collaboration • Worked with product teams, data engineers, and domain experts (orthodontists, clinicians) to align ML research with business needs. • Delivered explainable results for stakeholders without deep technical background.Patents & contributions • Co-inventor of a US patent (US20•••••••61A1) on ML/NLP in medical data analysis