A Nature Biotechnology perspective and accompanying work outlined integrated experimental and AI workflows for RNA structure determination, combining high‑throughput chemical probing with machine learning to overcome RNA’s dynamic and heterogeneous nature. The pieces argued for iterative loops between computational prediction and targeted experiments to resolve challenging structures. Industry actors developing RNA therapeutics can use these approaches to accelerate tertiary structure-informed design, improve target selection, and reduce downstream failure risk by validating predicted conformations experimentally.