Human-directed AI systems | Local-first agent infrastructure | AI literacy and assessment design
Tags: professional-portfolio, backend-systems, local-first-ai, education-technology, assessment-design, research-methods, live-facilitation
I build tools and learning materials where people stay authoritative, evidence stays visible, and AI supports cognition instead of replacing it.
| Core signal | Evidence |
|---|---|
| Backend systems building | Built Willow from scratch: Postgres data modeling, hybrid retrieval, task queues, MCP tooling, worker isolation, Grove messaging, and release discipline |
| Honest-systems engineering | Consolidated a multi-repo fleet into one core (one random stream, one tamper-evident record, one human-only seal) without changing behaviour; drove four interdependent PRs green across three GitHub organisations and shipped two automated releases (Case Study 4) |
| Sociotechnical research | Wrote The Illusion of Sovereign AI and maintained broader research vectors across AI sovereignty, institutional lineage, mechanics, and restoration |
| Education and governance | Designed AI literacy lessons and an assessment visibility framework for real classrooms |
| Live facilitation | Quizmaster for Geeks Who Drink since 2008: public speaking, pacing, audience trust, and real-time room management |
| Learning velocity | Moved from limited conventional SWE background into operating a multi-repo local-first agent platform |
| Public contribution | Emerging Rule lesson materials, SAFE app ecosystem, Willow public repos |
This folder is a technical-employer-facing packet. It translates work from Dispatches From Reality and related public projects into case studies a hiring manager can evaluate without reading twenty lesson files first.
| If you have... | Read this |
|---|---|
| 3 minutes | portfolio-case-studies.md |
| 10 minutes | willow-systems-portfolio.md |
| The newest work (2026-10) | one-script-fleet-case-study.md |
| Need research context | Research portfolio |
| Need systems evidence | willow-ecosystem-inventory.md |
| Need structure | case-study-template.md |
What this packet is: curated evidence of judgment, writing, systems-building, and design under real constraints.
What it is not: a résumé, a deployment report, or a claim that every artifact has been field-validated at scale. Where work is a framework, contributor material, or public prototype, the case studies say so.
| Document | Purpose |
|---|---|
| portfolio-case-studies.md | Three polished case studies for technical employers, including the Willow systems-building story |
| willow-systems-portfolio.md | Deep Willow case study: architecture, 10 proof points, claim boundaries |
| one-script-fleet-case-study.md | Case Study 4: the one-script consolidation, cross-repo CI, mutation testing, the honest forecast ledger (WIP) |
| willow-ecosystem-inventory.md | Curated repo census, SAFE apps, data artifacts, research substrate |
| case-study-template.md | Reusable structure for future writeups |
| Area | Entry point |
|---|---|
| AI literacy curriculum (HS) | lessons/ai-literacy-9-12-index.md |
| Assessment & governance framework | education/assessment-visibility-v1.1/white-paper.md |
| Research portfolio | research/README.md |
| Sociotechnical AI paper | research/sociotechnical-ai-sovereignty/README.md |
| K–12 story-first AI lesson | lessons/cs-k12-the-scribe-who-forgot-his-dreams.md |
| Teacher reading list | education/emerging-rule/ai-education-reading-list.md |
| Local-first agent infrastructure (current) | willow-memory org: willow-bot, willows-grove, willow-mcp |
| Human-verification ledger | Nestor |
| Local-first agent infrastructure (predecessor, archived) | willow-2.0 |
| Area | Detail |
|---|---|
| Operations | 27 years in grocery: store operations, inventory and ordering, scheduling, and keeping a floor running for the people on it |
| Education | Master's in Music Education; a year teaching middle school music; classroom materials written for working teachers (30 students, limited prep, devices optional) |
| Live facilitation | Quizmaster for Geeks Who Drink since 2008 |
| AI systems | Built Willow from scratch: local-first memory, task queue, knowledge graph, MCP tooling, Grove messaging, and verification loops |
| AI/ML credential | Hugging Face certificate |
| Contributor | Emerging-Rule/community: reading list, lesson proposals, AI literacy series examples |
All materials in this repo are CC BY 4.0 unless noted.
Portland, OR · open to remote · GitHub: rudi193-cmd · Portfolio writing: Dispatches From Reality