Engee documentation

What is it Engee MIND

Engineering development with an AI agent is a way of working in which an engineer sets a goal, initial data and acceptance criteria, and an AI agent plans actions, performs them in an engineering environment and checks the result. If the criteria are not met, the agent clarifies the decision and repeats the calculation. The engineer evaluates the assumptions and makes a technical decision.

This approach reduces manual operations when assembling models, changing parameters, performing a series of calculations and preparing reports. The checks can be repeated after each change, and the team’s methods can be saved as skills. The result depends on the completeness of the task, the quality of the data, the selected language model, and the checks used.

Purpose Engee MIND

Engee MIND (Modeling Intelligence for Nextgen Design) icon engee mind — AI module-automation of engineering development in the environment Engee. It connects the AI agent with the models, calculations, program code, data and libraries of the project.

The Engee MIND includes an MCP server. MCP (Model Context Protocol) — the protocol by which the agent learns the available tools and calls them with the specified parameters. MCP is a technical connection mechanism; Engee MIND is a built—in module that provides engineering tools, instructions and skills.

How the interaction works

engee mind about 1

The Language Model (Large Language Model, LLM) generates a plan and requests for tools. The agent client organizes calls and returns the results of these calls to the model. Engee performs engineering calculations. Models and saved files remain in the production environment Engee when changing a client or LLM; the transfer of chat history between clients is not implied.

What can be assigned to an agent

  • Create or modify the model, run the calculation, find the cause of the error and check the correction.

  • Develop an RF component for documentation and measurements using EngeeRF.

  • Transfer the model or algorithm and compare the result with the reference.

  • Create custom block or application Engee.

  • Connect local resources via Engee.Integrations and perform the specified operations with KPM RITM.

  • Convert a repeatable engineering procedure to custom skill.

Cloud and local work

For cloud Engee the address of the MCP server is used (accessible only from the AI agent). The local LLM can be connected via a compatible agent client. By itself, the local launch of LLM does not make the entire system autonomous: when accessing the cloud Engee the data is still transferred outside the computer. To work within the enterprise infrastructure, coordinated local components are needed, including Engee Circuit and the MCP server available on this network.

Limits of applicability of the result

The successful response of the agent does not yet confirm the correctness of the engineering solution. Require a saved model, actual run data, calculation conditions, and comparison with criteria. Separately check the applicability of physical assumptions and initial parameters.