Geminos Causeway is an end-to-end low code platform for modeling, building and deploying Causal AI applications.
Bridging the gap between data scientists and subject matter experts by providing
a common ‘visual’ language to collaborate.
Geminos Causeway is a robust, low-code platform for creating Causal AI apps and Causal Digital Twins. Built on Open-Source, Causeway provides a streamlined solution for causal and knowledge modeling with advanced causal analysis tools, including outlier, intervention, and root-cause analysis. ETL and data wrangeling is accelerated with over 4,000 integrations and plugins ready to drag-and-drop onto your canvas. Causeway brings versatility and efficiency to teams and organizations investing in Causal AI.
Visual Causal Modeling
The causal modeling feature of the platform is a powerful tool for visualizing and understanding the complex relationships between different variables in a problem domain. Providing a clear and concise representation of complex systems, users can create, collaborate and share models with the wider team or the rest of the business.
Causal Discovery with Chat-GPT
Leveraging the enormous knowledge-base and subject matter expertise available through Chat-GPT, Geminos has developed a causal discovery tool that identifies previously unrecognized causal variables and relationships while also constructing causal models for users. This significantly accelerates causal modeling, a crucial component in Causal AI and Causal Digital Twin solution development.
Causal Model Validation
Being confident that the assumptions captured by subject matter experts and the GPT discovery assistant are correct is critical. This short video introduces some of the concepts and techniques used to confirm or refute causal relationships and identify missing causal variables from your models.
Bias Detection and Removal
The Causeway platform has built-in tools to perform front and backdoor path analysis which supports bias identification within your causal models. Users can condition on the variables that act as colliders on backdoor paths to prevent spurious associations and eradicate bias which ultimately delivers more ethical AI. This can all be achieved visually without having to understand the causal math behind the analysis.
Knowledge modeling allows for the understanding and capture of your static data in a structured way. By mapping this knowledge to causal models, you can develop a more comprehensive and accurate understanding of the relationships between variables and the underlying causal mechanisms.
ETL and Data Hydration
The platform enables hydration of the causal models with data mapped through knowledge modeling so that causal analysis can be peformed. Over 4,000 ETL (extract, transform, load) functions and plugins are available as standard (e.g. CSV, SQL Server, Snowflake, RESTful API, etc.) which allows for faster data integration, reduced development time, and easier maintenance of data pipelines.
Intervention analysis, allows you to simulate the effects of hypothetical interventions on a causal model. This is a powerful tool in causal modeling because it enables researchers and practitioners to test different scenarios and explore the potential consequences of different policy or decision-making options, estimate the magnitude of effects, and evaluate the potential outcomes of different interventions.
More Videos Coming Soon