ML Engineer · Production AI Systems for Industrial Automation
ML engineer building production AI and automation systems for real industrial workflows. I ship document parsing, retrieval, LLM-assisted analysis, real-time file monitoring, and human-in-the-loop automation that runs under real-world constraints across Windows, Linux, and cloud.
- McQueenyML · Production HVAC estimating platform running daily on McQueeny Group's Windows servers. 5 manufacturer parsers, 195 scope templates, 419 tracked projects, an Ollama lifecycle analyzer, and a real-time R:\ drive watcher.
- CloudBooks · 25-table multi-tenant Supabase SaaS with row-level security, Stripe feature-gated billing, BYOK AI insights, and QuickBooks Online OAuth2 sync. Live on Cloudflare Pages.
- adze-cad · Native AI assistant for SOLIDWORKS: an in-process COM add-in with an agentic loop and governed write tools. MIT licensed, v0.1.1 public beta.
- Project Ava · Personal AI operations hub with durable session continuity, agent orchestration, and a template framework across ten downstream projects. Tailscale-only.
- cortex-design · Local-first learning system fusing Obsidian authoring, FSRS spaced repetition, and ChromaDB semantic search over a curriculum engine.
- adze-fusion · AI integration with Autodesk Fusion 360 for the multi-CAD adze product line.
Python · FastAPI · TypeScript · React · C# / .NET COM interop · C/C++ · PostgreSQL / pgvector · Supabase · PyTorch · Docker · Tailscale
Focus: applied LLM systems, computer vision (YOLO), CAD automation, multi-tenant SaaS.