Astromech is developing an AI platform designed to help researchers understand how genetic information gives rise to biological traits and systems. Its ambition is to build a “Large Life Model,” a biological intelligence system that combines deep learning with evolutionary analysis to interpret data across species and time. Rather than functioning as a conventional language model, the planned system is intended to work with multi-species inputs and transcriptomic data, using ancestral reconstruction to map evolutionary trajectories and investigate the causes of complex biological outcomes.

Astromech’s technical approach brings together attention-based learning for transcriptome analysis, Bayesian methods for ancestral reconstruction, graph-based pangenome modeling, and profile hidden Markov model alignment. A multi-omics integration component is also in development. The platform is intended to support research into therapeutic development, trait optimization, and programmable biology, with potential applications across medicine, agriculture, and broader life sciences. Astromech describes its system as under development, with the longer-term goal of moving beyond prediction toward tools that can navigate biological complexity and help researchers generate higher-resolution insights.