Researchers introduced MULTI‑evolve, an ML‑guided, lab‑in‑the‑loop platform that accelerates the directed evolution of multi‑mutant proteins by combining protein language models, targeted double‑mutant data, and iterative experimental screening. The approach compresses traditional multi‑round evolution into fewer cycles while capturing epistatic interactions crucial for high‑fitness variants. Complementary reports described a general lab‑in‑loop framework that integrates limited experimental data with models to explore high‑dimensional protein sequence spaces efficiently. Together, these methods aim to cut time and resource costs for protein engineering campaigns, with implications for therapeutic antibodies, enzymes, and synthetic biology.