I spend most of my time turning rough design ideas and raw inputs into inspectable, testable artifacts: OpenSCAD mechanism studies, fit-coupon kits, STL inspection utilities, local transcription pipelines, and reviewable render packages.

• safe mechanical prototyping
• OpenSCAD-based iteration for wearable and bench-top parts
• STL tooling and inspection utilities
• local transcription and interview-organization tooling
• AI workflows that still need real validation

• Parametric OpenSCAD prototypes for low-force wearable and bench-top mechanisms, with print validation and fit-coupon iteration
• STL inspection utilities for bounding boxes, triangle counts, surface area, and mesh metadata
• Reusable fit-coupon patterns for fastener sizing, slot testing, and clearance tuning before full prints
• Local transcription pipelines that convert customer interview recordings into transcripts, Q/A pairs, and structured research fields
• Render packages and part-organization workflows for turning rough CAD into something reviewable

• OpenAI Codex for implementation, refactors, review passes, and repo cleanup
• Prompt-driven CAD iteration support alongside OpenSCAD, slicer checks, and manual geometry validation
• N8N Workflows for optimizing workflows, and Obsidian for quick fixes
• Codex, Cursor, n8n, Google AI Studio, OpenSCAD, Vercel, Azure, Notion APIs, and local-first automation workflows.

• openscad-mechanical-prototyping: OpenSCAD-based wearable and bench-mechanism prototypes, render sheets, and printability notes
• stl-inspector-cli: Python CLI for inspecting STL bounds, triangle counts, and geometry metadata
• openscad-fit-coupon-kit: reusable OpenSCAD fit-coupon files for hardware, slots, and clearance testing
• transcription-organizer-pipeline: local recording-to-transcript pipeline with deterministic analysis, Q/A parsing, and optional Notion publishing