Organic synthesis is often a challenge that could slow innovation and increase risks and costs of research and development in the chemical industry. Modern machine-learning approaches for synthesis planning attempt to lower barriers associated with synthesis in discovery and process development. A recently released capability in SciFinder-n allows chemists to explore a vast space of synthetic routes for target molecules, defined by both known and predicted reactions. The predictive retrosynthetic search is powered by CAS's most comprehensive and accurate reaction collection, and is fully embedded in the SciFinder-n platform, allowing detailed analysis of the results and their application. In this presentation we will demonstrate how the tool can boost the creativity and productivity of chemists, and how with the thorough coverage of synthetic opportunities it can accelerate and improve the success rate of the synthesis of known and novel compounds.