Reaction optimization has historically been done from a “cold-start”, using a trial-and-error approach uninformed by historical data. For reaction optimization, we leveraged a method, known as multi-task Bayesian Optimization (MTBO), which learns from both historical reaction data as well as from new experiments. MTBO identifies optimal reaction conditions far faster than standard (single-task) Bayesian Optimization (STBO), which is not data-informed. Reactwise’s integration with AI4Green’s ELN creates a direct connection between experimental work in the lab and data-driven optimization. By linking Reactwise with the electronic lab notebook (ELN) used by chemists, the integration enables a smooth flow of information between the two systems. This helps researchers move beyond disconnected tools and creates a more unified workflow, where experimental knowledge and optimization can continuously inform one another.
At Reactwise, we offer full API integrations with external software/hardware platforms, enabling seamless data exchange across the digital laboratory ecosystem. This capability has been leveraged in our integration with AI4Green, allowing information to move efficiently between the ELN and the Reactwise platform. Through this two-way connection, historical experimental data stored in the ELN can be accessed by Reactwise to support MemoryBO and improve future recommendations, while experimental data generated through Reactwise, including information captured through PAT-connected workflows, can be transferred back into the ELN. The result is a flexible integration framework that reduces manual data handling, improves data continuity, and connects optimization directly with laboratory records.
Together, Reactwise and AI4Green provide a powerful foundation for more intelligent and efficient chemical development. AI4Green acts as a central environment for capturing and storing experimental data, while Reactwise uses that data to support optimization and decision-making. When combined with PAT and self-driving laboratory workflows, this integration helps teams make better-informed decisions, build richer experimental data sets, and accelerate research through a more connected and responsive digital lab environment.
AI4Green is dedicated to advancing sustainable chemistry through cutting-edge, data-driven methods, open-source innovation, and the promotion of FAIR data principles, developed by Professor Jonathan Hirst's research group at the University of Nottingham.
Learn more
