Machine Learning & Generative AI
AI, RAGs nemecky
Deep Learning for Structural Biology
I develop and deploy neural network models to solve complex categorization tasks, such as predicting the conformations of Complementarity Determining Regions (CDRs) in monoclonal antibodies. My work involves unsupervised representation learning and ensemble methods to map structural variations to functional outcomes.
Agentic RAG & LLM Orchestration
I design multi-agent systems using frameworks like LangChain and LangGraph to perform "Knowledge Mining." By building Agentic RAG (Retrieval-Augmented Generation) pipelines, I enable automated parsing of vast scientific literatures, transforming unstructured text into structured databases.
MLOps & Model Lifecycle
I implement modern MLOps practices to ensure model reliability, including automated retraining pipelines, hyperparameter optimization with MLflow, and monitoring for model drift.