Kaden VanHoecke
ML Engineer — BCI & Real-Time Systems

Kaden
VanHoecke

ML Engineer | Production AI Systems, Applied ML, and Industrial Automation

$
✉ Get in touch View work
Background

About

I build production AI systems that automate expert workflows. Not demos, not prototypes. Systems that run daily under real operational constraints.

I'm currently ML Engineer at McQueeny Group, where I built McQueenyML: a 74K LOC platform that cut HVAC estimating from 4 hours to 12 minutes with 100% parsing accuracy across 553K indexed files. I also built Adze, a native AI assistant embedded directly inside SOLIDWORKS with 19 grounding tools that read live CAD sessions. And Project Ava is a persistent multi-agent platform with structured memory, orchestrated domain agents, and 900+ tests across the stack.

I come from engineering coursework and the trades, not a CS degree. Everything I claim is shipped and quantified.

Now building toward real-time signal processing and edge ML. Currently developing SPDRbot, a quadruped robot on Jetson hardware, to build the closed-loop control and latency intuition that brain-computer interface work demands.

Selected Work

Projects

Project Ava

React 19
built

Dedicated headless server running as a persistent, self-evolving AI assistant. brain.db provides structured memory across 130+ sessions. 17 domain-specific agents handle tasks from tax compliance to 3D print calibration. Mobile-first PWA serves 15 feature tabs with inline editing, real-time monitoring, and Mercury banking integration. Express 5 API with Zod validation, ChromaDB semantic search, and OpenClaw agent orchestration — all accessed over Tailscale mesh from any device.

262
smoke-tested endpoin
React 19 Vite 8 Tailwind 3.4 Express 5 TypeScript 5.9 SQLite (better-sqlite3) ChromaDB OpenClaw
github →

CloudBooks

React 18
live

CloudBooks is a task tracking and business management platform built for informal employers — property owners, parents, and anyone managing real work without a formal business entity. React 18 + TypeScript frontend with Tailwind CSS, backed by Supabase Cloud (Auth, PostgreSQL with Row Level Security on all 22 tables, Realtime subscriptions, 6 Edge Functions). Features include hour logging with timer, multi-property workspaces with role-based access, payment processing, Pro-tier invoicing with PDF export, budget tracking, and AI-powered insights via OpenRouter/OpenAI/Anthropic BYOK integration. Stripe handles subscription billing with feature gating enforced at the database level. 288 unit tests + 61 E2E tests, GitHub Actions CI/CD, PWA installable with offline fallback. Security-hardened with CSP headers, JWT-authenticated Edge Functions, restricted CORS, and HTML-escaped email templates.

React 18 TypeScript Tailwind CSS Vite Zustand TanStack Query React Router v6 Supabase Auth
live →

McQueenyML

Frontend: React
built

Production automation platform for an HVAC contractor's sales engineering workflow. 5-tier matching engine (pgvector embeddings, trigram fuzzy search, Claude AI fallback) indexes 553K files and 3,000+ equipment orders synced from SharePoint. Processes 10K+ XML specifications with 100% parsing accuracy. Built on Next.js, FastAPI, PostgreSQL, and ChromaDB with real-time sync pipelines.

4h → 12m
speedup
Frontend: React TypeScript Vite Tailwind CSS. Backend: Python 3.11 FastAPI uvicorn. Database: PostgreSQL via Supabase (pg_trgm pgvector). ML: Ollama (qwen3:14b nomic-embed-text). Infrastructure: 7 NSSM services on Windows 11

Adze — AI for SOLIDWORKS

C# .NET 4.8
built

AI design assistant that runs inside SOLIDWORKS as a COM add-in, not alongside it. 19 typed grounding tools read live CAD sessions (dimensions, mates, mass properties, BOM). 8-step governed safety lifecycle for write operations ensures AI modifications are reversible and auditable. Built in C# on .NET with the SOLIDWORKS API. SOLIDWORKS Solution Partner Program application in progress.

C# .NET 4.8 SOLIDWORKS COM API NUnit 3 MSBuild PowerShell WinForms WebBrowser
github →

Oracle (Trade Signal)

Python 3.13 (miniconda3)
built

Self-directed market analysis system that generates hypotheses, tests them against historical data, and manages paper portfolios with real risk controls. 4 ML cortexes (LSTM prediction, sentiment analysis, technical patterns, anomaly detection) feed an Oracle decision engine with 8-dimension risk scoring. Walk-forward validation prevents lookahead bias. Built on Python, PyTorch, and FastAPI with real-time data ingestion.

8
risk dimensions
Python 3.13 (miniconda3) FastAPI SQLite TensorFlow PyTorch transformers (FinBERT) React/TypeScript/Tailwind (ava_hub UI)
Also Building

Document Processing Pipeline

Early McQueeny automation system that parsed XML specifications and generated Excel outputs for HVAC estimating workflows. Proved the automation approach with 100% parsing accuracy across 10K+ documents and directly led to expansion into the full McQueenyML platform.

3D Printing

End-to-end 3D printing system for LEGO-compatible part production, from parametric geometry and automated slicing to direct printer control and telemetry monitoring. Includes a dimensional calibration workflow with measured compensation sweeps and profile locking targeting ±0.05 mm tolerance on clutch-critical geometry. Built as a hardware-adjacent manufacturing pipeline with resolved slicer profiles, automated exports, and production telemetry.

SPDRbot

Robotics platform in active development for building practical skill in low-latency control, hardware debugging, sensor integration, and on-device inference. Represents a deliberate move from software-only ML systems into physical autonomy and real-world control constraints.

Capabilities

Skills

Python 92%
TypeScript 92%
Agentic AI Systems 92%
Automation 92%
PyTorch 78%
Docker 78%
Computer Vision 78%
Real-Time Systems 55%

Also in the toolkit

Languages

C# C/C++ SQL

Frameworks

React Node.js Express FastAPI .NET / COM Interop YOLO TensorFlow

Tools

n8n SQLite PostgreSQL Supabase ChromaDB Git / GitHub Tailscale Vite SOLIDWORKS

Domains

Neural Networks CAD / Mechanical Engineering
Home Lab

Infrastructure

A real mesh of four machines running production workloads — not a diagram for show. Ava is the headless Ubuntu hub serving the PWA control surface and running multi-agent orchestration. Frank, Hero, and Zoe are Windows peers on the Tailscale mesh.

🐧
Ava
Ubuntu · Hub Server
🖥️
Frank
Windows · Office
💻
Hero
Windows · Home
💻
Zoe
Windows · Compute
🔗 Tailscale Mesh 🔄 Syncthing 🐳 Docker 📊 Prometheus
Résumé

Experience

Founder & ML Engineer

VHTech LLC · Lenexa, Kansas

Jan 2023 — Present
  • Founded ML engineering consultancy supporting production automation and applied AI systems
  • Built and operate McQueenyML, a production HVAC estimating platform that reduced workflows from 4 hours to 12 minutes across 553K indexed files
  • Built Adze, a native SOLIDWORKS AI add-in with 19 typed grounding tools and governed write safety
  • Built Project Ava, a persistent multi-agent infrastructure platform with structured memory, private remote access, and operational monitoring
  • Built CloudBooks, a multi-tenant SaaS platform with real-time sync, subscription billing, and database-enforced access controls
  • Designed automation pipelines that parsed 10K+ XML specifications with 100% accuracy and proved the business case for full ML workflow automation
Python TypeScript C# React Node.js FastAPI Express .NET

ML Engineer (Contract)

McQueeny Group Inc. · Lenexa, KS (Remote)

Dec 2025 — Present
  • Sole developer of McQueenyML, a production HVAC estimating automation platform used by a sales engineering team
  • Built a 5-tier product matching pipeline combining pgvector embeddings, trigram search, and Claude fallback with a self-improving corrections loop
  • Reduced estimating workflows from 4 hours to 12 minutes while maintaining 100% XML parsing accuracy across 10K+ specifications
  • Automated SharePoint and Microsoft Graph ingestion across 372+ active projects and 3,000+ equipment orders
  • Operate a 7-service Windows production stack and maintain a 553K-row file index with real-time project lifecycle tracking
Python React TypeScript FastAPI PostgreSQL pgvector XML Excel

Sprayers Technician

Wallboard Specialties · Lenexa, KS

Jun 2018 — May 2025
  • Operated spraying equipment and handled material mixing on commercial job sites
  • Worked across seasonal and full-time trade crews while maintaining production pace and job site coordination
  • Spent 7 years in physical trade work while independently transitioning into software engineering and ML

Freelance Developer & Automation Specialist

Self-Employed · Texas

Jan 2023 — Present
  • Built and deployed CloudBooks SaaS accounting platform serving production users
  • Designed n8n automation pipelines reducing document processing time by 95%
  • Developed Project Ava: self-evolving AI assistant with 193 API endpoints and 368 tests
  • Created real-time ML systems including YOLO-based computer vision applications
Python TypeScript React Node.js Docker n8n

Education

Mechanical Engineering Coursework

Kansas State University · Manhattan, Kansas

Aug 2021 — May 2023
Get In Touch

Let's Build Something

Targeting ML Engineer roles with real-time signal processing and BCI focus. Open to consulting on production AI systems and ambitious applied ML work.

✉ kaden@vhtech.me ⚡ GitHub 🔗 LinkedIn

📍 Lenexa, Kansas