Get started with out-of-the-box pre-built models that come pre-trained and ready for use. Capturing the most common predictive use-cases, these pre-built models help generate AI based insights for use in key workflows immediately.
The Medius AI Studio enables key business stakeholders such as analytics professionals to test the pre-built models in a controlled sandbox with minimum effort, and experiment with various data combinations. All pre-built models are served through standard APIs in production environments.
A score between 0 to 1000 as an indicator of a proposer's insurability risk on the basis of health and financial variables. For eg: a score above 800 indicates preferred class, a score below 400 indicates a sub-standard class.
A score between 0 to 100 as an indicator of an individual's health status based on their exposure to different health risk factors, medical history, family history, lifestyle, and nutrition. For eg: a score above 75 indicates an applicant in good health.
Extracts, decodes, and analyses lab test results from diagnostic reports in locked formats (such as images and pdf) using advanced object recognition techniques and natural language processing algorithms to automatically perform activities such as
- unstructured medical text and entity extraction
- extracting laboratory test names, values, units and reference intervals
- detecting the existence of meaningful relationships between multiple entities/types in unstructured text
- use abnormal lab values, correlations, and relationships to infer existing conditions, and predict disease risks
- predicting individual morbidity and mortality scores using lab test entities and their values