Neural Concept raises $100M to bring AI into CAD and simulation workflows
Neural Concept SA, a Swiss AI startup spun out of the Swiss Federal Institute of Technology in Lausanne, secured $100 million in Series C funding. The round was led by Growth Equity at Goldman Sachs Alternatives, with Forestay Capital, Alven, HTGF, D.E. Shaw Ventures and Aster Capital participating.
The company embeds deep learning into everyday CAD and physics-based simulation tools. The premise: models that "know" geometry, physics and design intent can surface performance and constraints early, so teams can move with more certainty and fewer reworks.
Why this matters for product development
Traditional engineering loops are slow: design, simulate, revise, repeat-then hit surprises at prototype. Months disappear in iteration, and late-stage changes burn budgets.
Neural Concept pushes analysis into the earliest steps. Teams can explore more design options in the same time window, see tradeoffs sooner and cut the risk of painful tweaks downstream.
- Early performance insight inside existing CAD/CAE workflows
- Faster design-space exploration with clear tradeoffs
- Fewer prototype surprises and reduced material waste
- AI that acts as an "engineering intelligence" and improves with data
What the platform is doing today
The software is in use across automotive, aerospace, energy and advanced manufacturing. Over the last 18 months, revenue grew more than 4x, and the company now serves 50+ global enterprises.
The team highlights examples like "F1 aerodynamics predicted in 0.3 seconds by deep learning," underscoring speedups that change how quickly concepts can be evaluated.
How team roles shift
As AI handles more of the heavy computation, engineers can invest their time in intent, risk, performance vs. cost, and final decisions. The tool influences how choices are framed and justified with data.
"We founded Neural Concept with the ambition to enable complete AI-driven design of advanced systems like tomorrow's cars and spacecrafts," said Founder and CEO Pierre Baqué. "Advances in AI are transforming engineering from a process of trial and error into a data-driven workflow where tradeoffs and constraints can be understood and optimized from the start."
What's next
The new funding will accelerate platform development, including a generative CAD feature planned for early next year. The company is also expanding its go-to-market teams and deepening integrations with existing design and simulation tools.
Practical steps for product leaders
- Pick a high-iteration subsystem (e.g., aerodynamics, thermal, topology) and run a pilot with AI-in-the-loop to benchmark cycle time, cost and defect rates.
- Lock down data standards: units, meshes, materials, naming, versioning. Clean data turns into better models and fewer regressions.
- Add guardrails: physics-based checks, surrogate model validation, and pass/fail criteria before any AI-suggested geometry moves forward.
- Integrate where engineers already work-inside your CAD/CAE stack-so the process improves without extra friction.
- Track the right metrics: iteration count, lead time, design-space coverage, number of late-stage changes and prototype defects.
- Upskill your team on AI-assisted design workflows and verification methods. If you need structured paths by role, see AI courses by job.
Context and sources
- Origin: project from the Swiss Federal Institute of Technology in Lausanne (EPFL).
- Lead investor: Growth Equity at Goldman Sachs Alternatives.
Your membership also unlocks:
