Neural Concept launches AI tool for engineering software

Neural Concept has launched MCP Bridge, a tool that lets AI agents operate commercial engineering software. It is already in use at companies including OPmobility and TRATON.

The launch extends the Swiss company's engineering platform into the computer-aided design, model preparation and simulation workflows used by manufacturers. MCP Bridge connects AI agents with software including CATIA and NX for design, ANSA and HyperMesh for model preparation, and Abaqus for simulation.

The software is intended for corporate environments where engineering teams work across different systems, licensing arrangements and security requirements. Built on the open Model Context Protocol, MCP Bridge is designed to run across Windows workstations and Linux clusters.

Manufacturing groups have been testing how AI can be used in engineering beyond code generation and document handling. In product development, much of the manual effort still sits around the simulation process, including finding the right geometry, cleaning it, preparing meshes, setting up the physics and returning results to the design stage.

MCP Bridge lets AI agents handle those steps while engineers define the objective and review the outcome. Customers retain control over their programs and data.

In one customer example, an engineer using MCP Bridge cleaned up the geometry of a wheel rim in about three minutes, compared with a day using a manual process.

Pierre Baqué, Chief Executive Officer and Co-Founder of Neural Concept, said the product marks a shift in how AI can be applied to industrial software. "For the first time, the software that shapes every physical product can be operated by AI at enterprise scale. The real opportunity is to give engineers more leverage, so they can take on harder problems and solve them faster," Baqué said.

Tool access

The approach combines software connectors with task-specific instructions for how tools are used in engineering settings. Those connectors reflect each team's internal rules and established ways of working, with the aim of fitting AI agents into current processes rather than requiring engineers to change them.

That focus reflects the fragmented nature of engineering software. Product development often moves through several specialist applications, with different file formats and workflows across design, model preparation and simulation. Companies in sectors including automotive, aerospace, electronics and energy often rely on a mix of commercial software running on local machines and central computing infrastructure.

Neural Concept benchmarks AI models on engineering tasks such as CAD editing and meshing, then assigns the most suitable model to each step. This reflects the fact that no single model performs best across all stages of an engineering workflow.

Customer use

Early users include engineers at TRATON and OPmobility, both of which are exploring how AI agents can work with existing engineering toolchains.

"AI agents have completely changed the way I work. I can run more work in parallel, and going from a research paper to a validated script in our simulation toolchain now takes a couple of days. The next-generation engineer focuses on planning and evaluation, managing teams of AI agents working towards a goal. When agents can operate the engineering tools we already use, nothing is out of reach anymore," said Ciro Alosa, Electromagnetic Design Engineer at TRATON and Neural Concept user.

OPmobility described a similar line of inquiry, with an emphasis on keeping engineers in control of design choices.

"At OPmobility, we are exploring how AI agents can operate the CAD and simulation tools our engineers already use. The aim is to give engineers more time to explore design options, while they keep full control of the decisions that require engineering judgment," said Yoann Eulalie, R&I Group Leader at OPmobility.

The product is available to Neural Concept customers as part of what it calls its Engineering Intelligence Layer. The next step is for agents not only to operate a company's software tools but also to learn how that company designs.

Founded in Switzerland in 2019, the business serves customers across automotive, aerospace, energy, consumer electronics, semiconductors and defence, and has expanded beyond Europe into the US, South Korea and India.