Wi!Mi (KESMI): environment for designing knowledge models

Accumulation, dissemination and transmission of knowledge from generation to generation have always determined the development of human civilization.

Knowledge management systems are a necessity in the modern information world, and expert knowledge along with technology are key resources for company development.

Wi!Mi (KESMI – Mivar expert system designer) is a tool for designing knowledge models with unlimited number of connections, parameters and relations, which has logical inference.

Wi!Mi can be efficiently used for developing software robots (RPA) or virtual specialists, complex expert systems (ES), knowledge management systems (KMS) or logical reasoning systems (LRS).

The basis of Wi!Mi is an entirely new approach to description and formalization of any types of knowledge – mivar-based approach.

Mivar-based approach is a universal modelling system that allows us to use all the advantages and capabilities of available tools for working with knowledge efficiently such as ontologies, cognitive maps, ER-models and semantic networks.

The use of mivar logical inference with linear computational complexity allows us to process more than 5 000 000 rules in a second with minimum hardware requirements.

Practical application

  • Decision support systems (DSS)
  • Repair management systems
  • Logical module for situation centres
  • ERP intellectualization
  • Logical inference for BI-systems

Solution components

  • ENGINE performs the function of logical solution calculation using input parameters
  • GUI is an interface for creating, editing and testing models
  • API: Rest API allows us to communicate with Engine through HTTP using JSON format.

The tasks solved

  • Developing expert systems, knowledge management systems (KMS);
  • Designing logical reasoning systems;
  • Designing and managing knowledge models in any subject domains;
  • Generating execution algorithms: logical inference.

Functional capabilities

  • Visualizing mivar models in the form of a table/a list/ a graph;
  • Structural analysis of the models considering contradictions, correctness and input data completeness;
  • Testing models; Logical inference visualization.


  • The use of available knowledge for processing;
  • Automated generation of solution algorithms on the basis of available knowledge.

Supported platforms

  • Windows 7, 8.x, 10
  • OS X from 10.9 to 10.14
  • Linux