CAS, a division of the American Chemical Society specializing in scientific knowledge management, announces the launch of the next evolution of CAS SciFinder, the leading scientific search and insights platform. By integrating advanced AI-optimized for scientific applications-with the largest human-curated collection of global scientific knowledge, CAS SciFinder empowers researchers across R D to work faster and unlock new opportunities for innovation.
This marks the most significant integration of AI within the CAS solution portfolio to date. Scientific information is diverse and complex, making it challenging for common AI tools to interpret accurately, noted Tim Wahlberg , CAS Chief Product Officer. Applying these technologies to the high-quality, structured data in the CAS Content Collection and engaging our team of scientists to verify the advanced AI models enabled us to develop a science-smart AI approach that delivers accurate and reliable answers scientists can trust.
SearchSense, a collection of new AI-driven search capabilities in CAS SciFinder, was developed to make scientific information more accessible across R D workflows and help researchers reach actionable answers faster. With a single search box, users can initiate a broad range of inquiries by asking simple questions using natural language. AI-enabled summaries help users quickly interpret results and pinpoint the most relevant information to provide a clear answer they trust. Ninety-three percent of beta testers agree that the new search features in CAS SciFinder make them more efficient. The platform also learns from user activity to personalize results and streamline future searches. To ensure the security of all user queries, results, and interactions in CAS SciFinder, AI features operate within a closed system.
CAS also continues to lead innovation in retrosynthesis with the introduction of the first real-time, interactive synthetic planning tool. This is a paradigm shift for chemists that slashes the time needed to generate a synthetic plan from minutes to just a few seconds. By eliminating wait times and providing real-world evidence directly in synthetic maps for quick comparison, Interactive Retrosynthesis in CAS SciFinder significantly increases efficiency and enables greater creativity in reaction planning and optimization. Scientist-enhanced AI is also being leveraged for more accurate interpretation of users reaction queries to deliver unmatched search accuracy and precision within the largest collection of curated reaction information.
To address a growing need among scientists and stakeholders in all phases of R D to have greater visibility of the intellectual property IP landscape, additional IP search support, visualization capabilities, and patent details have also been added to CAS SciFinder. Users can now easily identify prior art within their typical workflow with the help of IP Connections. This CAS-developed AI algorithm matches relevant patent and non-patent publications from any free-text input to inform more confident prioritization and reduce the risk of wasted time and investment.