SolidWorks introduces conversational AI assistants to speed up CAD work
At 3DEXPERIENCE World 2026 in Houston, Manish Kumar, CEO of SolidWorks, outlined how generative AI is being woven directly into the CAD workflow. The pitch is simple: talk to SolidWorks like you would to a chatbot and watch complex geometry, documentation, and checks come together faster.
SolidWorks, a core product from Dassault Systèmes with a large share of the global CAD market, is pushing AI across design, drawing, simulation, and 3D modeling. The goal is to lower the barrier for design without taking control away from engineers.
Meet the trio: Aura, Leo, and Marie
Aura is the context coordinator. It keeps track of project intent, remembers prior decisions, and helps you explore ideas in natural language. Think of it as the conversation layer that knows your design history and objectives.
Leo is the engineering checker. Share a concept and Leo evaluates feasibility, flags constraints, and points out what will or won't build.
Marie is the science brain. It provides guidance on materials, chemistry, and microbiology when your design crosses into those domains.
Example: ask, "What material should we use for the wing of this electric hydrofoil?" Aura surfaces viable options, Marie explains the scientific trade-offs, and Leo confirms whether a chosen material can be manufactured for your geometry and loads.
Chat-driven design, start to finish
You can now drive modeling with instructions like: "Create a shape 100 cm wide and 20 cm long," "Select vertical edges and apply fillets," and "Drill two holes on the top face." The idea is fewer clicks through menus and more intent expressed in plain language.
According to the company, this flow reduces time and cost by cutting repetitive steps, automating setup, and reusing knowledge from existing designs and documents.
What this means for product teams
- Faster concept-to-geometry loops without hunting through toolbars.
- Context carried across design, simulation, and drafting to limit rework.
- Knowledge reuse from legacy models, drawings, and specs.
- Broader participation: non-experts can propose design variants while engineers oversee quality and constraints.
Kumar put it clearly: AI is an engine; engineers still drive. It can automate repetitive work, but it doesn't truly grasp the physical world-human judgment remains essential.
How to prepare your workflow
- Clean your data. Organize parts libraries, materials, and drawings-these are the sources the assistants will rely on.
- Set guardrails. Lock in standards for tolerances, materials, safety factors, and release criteria so AI suggestions stay within your rules.
- Start small. Use AI for fillets, hole patterns, simple extrusions, drafting, and documentation before handing it complex assemblies.
- Validate by default. Pair AI-generated geometry with simulation, DFM checks, and peer review before release.
- Upskill the team. Teach prompt patterns that yield consistent results and how to review AI outputs with a critical eye.
For event context and future updates, see Dassault Systèmes' 3DEXPERIENCE World page: Official event site. SolidWorks product details are available here: solidworks.com.
If you're planning team training around prompts, QA, and review cycles, explore practical courses here: AI courses by job.
Your membership also unlocks:
