Mitchell Dawkins
Real products, honest lessons, live. The projects and journal are the record: architecture, mistakes, and what shipped. I also help lead the Dawkins Family Foundation.
Consulting is open for custom software, AI and business automation, and modernization.
Project stories covering problems, architecture, and lessons. Not just finished screenshots.
A video grid in one window: up to 16 streams, one audible channel, scorebugs, close-game alerts, and a same-Wi-Fi phone keypad.
A static browser app of early-learning games. One learner record in localStorage drives every round. Questions come from a CSV. Speech and pose tracking stay on the device.
An offline coaching-film player and a Hudl-scoped Chrome extension: download the clip in the browser you are already logged into, then review it with timestamped telestration, notes, and play data when the network is gone.
A static practice ballot for the Grand Rapids area. One county contest list, filtered by precinct scope, with millage questions priced from a taxable value. Marks stay in the browser.
An offline-first biblical fasting companion PWA for a 190-day journey — daily plans, scripture, check-ins, journal, streaks, badges, and optional Supabase groups with Web Push reminders.
An AI-native operating system for repositories that orchestrates planning, architecture, implementation, testing, documentation, deployment, and continuous learning through specialist engineering agents.
A microlearning PWA that replaces doomscrolling with a one-card feed of ideas, news, Gutenberg books, and free learning sites — plus catalog-grounded Ask via OpenAI or Ollama.
A self-hosted VPS platform for deploying and managing a fleet of WordPress sites, each running as an isolated Docker stack with Git-driven deploys, centralized MainWP management, automated backups, and a FastAPI + React control plane.
Multi-tenant AI system for processing video content, extracting insights, and training custom models
Advanced WNBA analytics platform with AI-powered predictions, prop betting scanner, Monte Carlo simulations, and comprehensive statistical analysis.
Advanced NBA analytics platform with AI-powered prop bet predictions, machine learning models, real-time ESPN data integration, and comprehensive betting analysis tools.
A fully self-hosted microservices stack that turns any article URL into an NPR-style podcast episode using local LLMs and zero-shot voice cloning text-to-speech.
A film-grading pipeline and coach hub for weekly and season-long receiver evaluations — year-two v2 adds drop-a-CSV ingest, a two-season landing page, and a player-facing rubric.
Articles from real builds: architecture write-ups, failures, and practical engineering.
I built Film Room so a cutup, the telestration, and the notes still work on a bus, a sideline, or a locker-room wifi that keeps dropping — and so a Hudl playlist does not have to wait on an email before it can be reviewed.
A long November ballot is a stack of names and millage rates. I built a practice page that keeps one precinct’s contests on a single sheet and turns the tax questions into dollars.
I built Video Grid so a full slate fits in one Chrome window: multiple streams, one audible channel, scores on the panel that is showing that game, and a phone keypad on the same Wi-Fi.
Senior Full Stack AI Engineer
People follow how I think, not a single app. Engineer, founder, teacher, and service provider: I build real products with AI, stream the process, and turn every lesson into something others can learn from. I also help lead the Dawkins Family Foundation, continuing a family legacy of service. Full resume and background live on About.