Oral Presentation
Investigating the Implementation of AI in Katalyst D2D
Alexander Waked; Solution Area Manager, Katalyst D2D; ACD/Labs
AI-enabled reaction design has long empowered scientists within Katalyst D2D through integrable third-party tools. The rapid evolution of artificial intelligence is opening new frontiers. We have investigated how large language models (LLMs), machine learning predictions, and emerging AI capabilities can help scientists accelerate experimental workflows, enhance decision-making, and extrapolate insights beyond observed results.
This presentation will share our findings across four key areas:
Streamlined experiment building — leveraging LLMs to design experiments with fewer clicks through intuitive text prompts
Assisted results interpretation — AI-supported analysis to help scientists draw faster, more confident conclusions from experimental data
Predictive extrapolation — applying transition state information, reaction kinetics, and property prediction to extend findings across broader chemical space
Automated reporting — intelligent generation of experiment summary reports to reduce manual effort and improve documentation consistency
Learn More