top of page
Minerva UI

Solving scientific inference problems in complex systems

Our modular scientific modelling framework, Minerva, is built on Bayesian probability and graphical models, and is used by some of the world's largest scientific experiments. It generalises modelling and runs your model in the cloud.

Product

A scientific modelling framework

Modularity

The Minerva framework can be used to build small or complex scientific models in a modular graphical fashion with reusable components, shared within and between organizations. Use cases are inference, predictive simulation, uncertainty quantification, experimental design and data fusion.

Standard for scientific models

Minerva defines a standard for describing scientific models, using either API, notebooks, or graphical tools. Standardised models can be easily interpreted and understood by anyone in the organisation, and freely exchanged, stored and versioned.

An analysis infrastructure

The Minerva cloud based modelling and analysis framework can automatize all your analysis for your organisation, running complex and simple models routinely at scale.

Cloud

Minerva models run as services on a scalable cloud with CLI access from your organisation, or installed on premise for organisations that manage their own cloud resources. 

Machine learning

Minerva can create surrogate machine learning models automatically from a model specification, giving orders of magnitude faster analysis of complex models, which can be used for e.g realtime analysis and control systems. 

Consulting/tutoring

We also do consulting work and tutoring in scientific modelling, analysis and experimental design. Using our framework and cloud infrastructure we can substantially reduce the time from developing of models to routine analysis at scale, using advanced Bayesian hierarchical models and machine learning, applied to your problem domain.

Accelerating scientific discovery

Our unique approach to building simple and complex scientifc models can be used for both predictive simulation from models, experimental design, and inference from observations

Nuclear Fusion

Minerva is used in the private and public nuclear fusion industry for analysis of the multitude of measurements from nuclear fusion experiments. Delivering deeper understanding of the internal structure of fusion plasmas from magnetic field measurements, x-rays, cyclotron emission, spectroscopy and other diagnostic systems.

About

About

Background

The founders of Seed eScience Research have a background in nuclear fusion research with special focus on advanced modelling and analysis, and have authored several hundred scientific papers. 

Innovation

A generic scientific modelling and analysis framework; the Minerva framework is designed to handle advanced modelling and scientific analysis on an unprecedented scale, being highly modular, using Bayesian probability theory, hierarchical models, machine learning and modern cloud technologies.

Vision

We believe that one of the greatest limitations to the pace of scientific progress today, is the lack of an organized and consistent approach to scientific modelling & analysis. We want to accelerate this process substantially, bringing scientific discovery to the next level.

Get a Demo

Get in contact

Schedule a meeting, we are excited about your project

THANKS FOR SUBMITTING

bottom of page