Fundamental agricultural research.
Glassbiome is an agricultural deep-tech company developing new capabilities in crop performance, useful compounds and resource-efficient production. Our approach integrates AI research, controlled cultivation and laboratory science.
Powered by Century AG v2
Century AG v2 coordinates supervised AI agents that investigate scientific evidence, compare mechanisms and challenge predictions in parallel. Our research programme connects those predictions with cultivation and laboratory experiments, designed to generate original data for model training, independent validation and the next cycle of discovery.
Our focus areas
Start with industry needs.
Investigate how biology could address them.
Beauty & personal care
How can biological compounds meet formulation and production requirements?
We investigate how growing conditions and processing affect compound composition, stability and function, then define the tests needed for a finished formulation.
Personal care · beauty · ingredient development
Food & nutrition
How do biological and processing choices affect the quality and usability of food?
We investigate how cultivation and processing shape nutritional composition, ingredient performance and production economics, with experiments designed around the intended food application.
Nutrition · functional ingredients · food systems
Agriculture & resources
How do cultivation conditions change useful output and resource demand?
We investigate responses to light, water, nutrients and other growing conditions, connecting crop yield and composition with resource use and production requirements.
Agriculture · resource efficiency · production
Circular industries
When can a byproduct become a useful input?
We examine recovery and processing routes for byproducts and underused biomass, considering usable output, quality, resource demands and competing uses.
Circular materials · biomanufacturing · recovery
The Century research engine
Supervised AI agents examine crops, compounds, growing conditions and processes across original studies to identify connections worth testing.
Check study design, measurement and contradictory results. Separate what a paper demonstrates from what a model infers.
Compare mechanisms, models and environmental schedules alongside usable output, resource demands and production constraints.
Design cultivation and laboratory experiments to test predictions and generate original data. Use measured responses, including failed predictions, to refine models and select the next experiment.
Scientific evidence, model predictions and experimental results have distinct roles. Each research cycle is structured to test assumptions, preserve contradictory results and identify what the next measurement should resolve.
Inside the platform
A view of the industries and questions
that shape our research.
Illustrative overview · research themes
Physical research & original data
Controlled cultivation is central to our research: an experimental system for studying how plants respond to changing environments. Our study designs connect conditions during growth with plant composition, recovered output and performance in the intended product or process.
Experiments are designed to produce reusable datasets: growing conditions linked to plant responses and laboratory measurements. These records provide the basis for training and testing predictive models, and for identifying production methods worth scaling.
Leadership
Co-founder & CEO
Co-founder & CTO
Research & commercial partnerships
We welcome research and industry partners who can help test and develop applications in crops, ingredients and biological materials.
Detailed findings, scientific theses and supporting analyses are shared through the private partner workspace.