We talk with Makistry founder Parth Meta about turning plain English prompts into parametric CAD and why a structured AI “brainstorm” can speed design without losing engineering control. We dig into standards-aware reasoning, exports, limitations, and the roadmap for assemblies and 2D-to-3D.

• Why CAD still slows real projects
• Text to parametric models through a guided brainstorm
• Using RAG to ground standards like M4 and hole specs
• Open Cascade kernel, STEP exports, in-browser visualization
• Measuring and parameter checks for trust and repeatability
• Limits on complexity and plans for sketches to 3D
• Assemblies, mates, and constraints on the roadmap
• Moving beyond copilots to AI-native design and CAM
• Balancing startup work with school and sport
• Making geometry organic yet manufacturable