Larger datasets and more flexible design for chemicals, materials, and life sciences R&D teams
Cambridge, UK – November 21st, 2023 – Intellegens today announced the 2023 Autumn Release of the Alchemite™ machine learning (ML) software, delivering more value from real experimental and process data for teams in chemicals, materials, and life sciences R&D. This release optimises usability for larger, denser datasets, delivering deeper insights in the race to develop improved formulations, materials, chemicals, and processes. It enables more flexible exploration of design space, broadening the applicability of tools that are proven to deliver a 50-80% reduction in experimental workloads when compared with conventional design of experiments (DOE) approaches. And it helps R&D teams to onboard new users faster and to collaborate based on analysis and guidance from ML predictions.
“The Alchemite™ 2023 Autumn Release summarises six months of work during which we have continually collaborated with users to prioritise, design, and deliver enhancements based on their real workflows,” said Ben Pellegrini, CEO of Intellegens. “We’re seeing the results in an expanding back-catalogue of successful projects and a growing user-base. Recent examples have included new or extended projects with R&D leaders in chemicals, foods, fast-moving consumer goods, biopharmaceuticals, alloy development, and plastics.”
New capabilities for larger, denser datasets include algorithmic improvements that deliver faster results for common experimental scenarios, a slicker user experience in interacting with this data, and extensions to the scope, sorting, and filtering features of key analytics when applied to thousands of rows of data.
More flexible and powerful exploration of design space is enabled by further algorithmic developments and usability improvements. Greater accuracy in predictions and in estimating of uncertainties are both crucial when using ML results to guide decision making. User experience enhancements make it easier to specify and analyse results from studies of experimental and process scenarios involving tens or hundreds of ingredients and processing variables with complex, overlapping constraints.
Finally, support for real-world R&D teams has been advanced through a host of small user experience and collaboration feature changes that ease onboarding of new team members and support quick reporting and sharing of results with colleagues, partners, and managers.
