La Agente Optima Framework Automates Bayesian Optimization in Self-Driving Labs

This framework separates LLM reasoning from experimental execution to supervise Bayesian optimization campaigns across computational and physical platforms. It maintains a persistent optimization state, returning control to the agent only when interpretation or campaign revision is required for scientific discovery.