Democratising real-world drug discovery through agentic AI

• Agentic systems are an emerging technology for automation, but evidence of their real-world use in drug discovery is limited. • We present AstraZeneca’s agentic systems LangDMTA and ChatInvent, outlining implementation challenges and lessons learned. ChatInvent, a conversational interface for agentic drug discovery, is illustrated with a typical use case. • Key technical considerations are discussed, including architecture choice, automated testing for continuous integration, and benchmarking. • Practical recommendations are provided for building agentic systems in drug discovery. • We speculate on the potential of agentic systems for improving efficiency in pharmaceutical R&D. Agentic systems that are based on large language models (LLMs) have emerged as promising tools in the chemistry domain over the past few years. Early examples included work on CoScientist, Chemcrow, and LLM-RDF, which showcased the potential of agentic systems to assist in chemical research, in the orchestration of cheminformatics tools, and in synthetic reaction development. Despite this, the current literature lacks examples of the real-world adoption of such systems in drug discovery. We present such an example by describing our work on an agentic system called ChatInvent, which has been integrated into the discovery pipeline at AstraZeneca to aid in molecular design and synthesis planning. We discuss how the system evolved from a proof-of-concept single agent into an extensible, robust, and scalable multi-agent architecture with a graphical user interface. We emphasize the lessons learnt and the challenges that persist as we continue to work on this project, and share our perspectives on the future of agentic systems in our domain.