MINERVA
Minerva is a platform for invoking and interacting with AI agents in scientific research. Built on our experience in bibliometrics, genomics, and cancer research, it lets researchers use agents to explore datasets, test hypotheses, and accelerate discovery.
DeepMind's co-scientist and similar systems work by churning away attempting to automate scientific thinking. Our perspective is that meaningful scientific advances require human researcher intuition, hypothesis refinement, and course correction throughout the process—not just at the end.
Minerva takes a different approach. Instead of autonomous agents that disappear for hours, researchers invoke and interact with agents step-by-step while using AI to handle complex research workflows.
Minerva helps researchers:
- →Navigate the research landscape and identify promising directions
- →Test hypotheses by simulating protein behaviour with AlphaFold
- →Spin up compute to train models for specific data types—like re-interpreting GWAS data with epistasis transformers or hybrid GNN architectures
- →Run parallel AI conversations that branch and merge while maintaining context awareness across the workbench
- →Collaborate with teams, sharing workflows and datasets
We're building Minerva for biotech, pharma, academic, and federal health organisations. We have early prototypes running and are actively expanding capabilities.
