Selected work

What EGI has shipped

Production systems, the framework underneath them, and the public libraries they feed. Everything on this page is running, public, or verifiable. Client engagements are covered by confidentiality and are described by shape, never by name.

By the numbers

From the GitHub ledger published with the CV, generated 2026-09-03, covering 14 repositories from December 2025 onward. One engineer, working solo, with agentic tooling.

2,902
commits across 14 repositories
1,176
reviewed pull requests merged across the same 14
4,648
automated backend tests in the flagship system and GWEDU, green on every merge
38
active sources in the CODEX discovery watchlist

Systems

Built and operated end to end: architecture, implementation, CI, deployment, and the on-call that follows.

Flagship · live

Workflowtech.ai live

Architecture · implementation · operations

A public AI terminal that is neither an open chatbot nor decoration: structured intake routes each visitor into a persona mode before the first model call, and the backend owns every decision after that. It runs as a live prompt session, bilingual in English and es-CO Spanish.

  • Local-first model routing with health-gated cloud escalation
  • Spend governance with per-turn, per-day, and dollar ceilings
  • An accounting breaker that treats unmeterable spend as an outage
  • Hybrid retrieval over the curated CODEX corpus, which is the archived Rewst automation library plus EGI reference material: an on-disk FTS5 lexical index fused with dense vectors, then reranked by a cross-encoder
  • Automated discovery pipeline feeding a public library and digest
Infrastructure · internal

LLM Builder Kit

Control plane · doctrine · 623 commits since 2026-06-11

The factory the rest of this page is built in. A control plane for running a one-person software studio at multi-team throughput: a native operator cockpit, a token-gated web panel behind it, and a worker fleet that keeps the queue moving while the operator is asleep.

  • An Avalonia desktop cockpit with fourteen workspaces over the same control-plane API the web panel uses
  • Thirteen registered worker loops covering routing, fulfilment, outreach, and the scheduled operator brief, four of them gating the panel's health check
  • A self-hosted LiteLLM key broker fronting the cloud Anthropic route, so rotating that credential is an operations task and never a deploy
  • One source of truth for engineering standards, vendored into the core stack with a drift check in each of those repositories' CI
  • Tunnel-only by construction: the panel binds to a WireGuard address, and its deploy script refuses a wildcard bind and tears down any public vhost it finds
Open source

GestaltWorkframe

Framework extraction · public repository

The terminal generalized into a reusable, brandable framework: FastAPI backend, Next.js terminal, and deployment bundles that make branding a directory of configuration instead of a fork. Extracted from the production system in May 2026 and developed in the open since.

Platform · in development

GWEDU in development

Architecture · implementation · 1,007 commits since 2026-05-28

A self-generating smart LMS. Name a subject or upload your own documents; the system researches it, bounds it in cited sources, designs a curriculum grounded in learning science, generates the materials and assessments, grades its own output, and publishes a branded learning portal that runs on your infrastructure. A learning system, not a quiz site.

  • Every generated claim is checked against the retrieved source text; unsupported items are rejected and regenerated, not just flagged
  • Pedagogy itself is grounded in cited education science, with continuous checks that its own evaluations stay fair
  • A domain-agnostic engine with snap-in Program Modules: content and configuration only, no engine code
  • A capability factory: when a course needs a tool the platform does not have, the platform builds it, wires it, and the lesson calls it
  • Practice environments: a course can hand a learner a real sandbox and declare what a stranger's shell is allowed to do in it
  • Self-hosted and air-gappable; modules install from signed bundles through a CLI that validates the manifest and enforces a framework-compatibility gate
  • Locale-aware delivery, tiered entitlements, multi-tenant host routing, xAPI completion records, and certification
Open source

CODEX

Curated public library · automated pipeline

A Git-backed catalog of community-built automation and GenAI-Ops content, assembled by a scheduled discovery pipeline and curated by explicit operator review. Human-browseable on the site and machine-readable for AI crawlers.

Open source

MSP Automation API

PowerShell library · observability and testing

A modular PowerShell automation library for MSP administration, built with the operational furniture such scripts usually skip: structured observability, security boundaries, and a test suite.

Open source

Claude Certification Exam Prep Coach

Assistant skill · public repository

The study loop used to earn the Anthropic certifications this page claims, released as a skill: profile, baseline diagnostic, a phased plan sized to a real calendar, and a persistent miss log. Three lanes under one coach: Architect Foundations, Architect Professional, Developer Foundations.

Open source

Job Hunt

Assistant skill + local dashboard · public repository

A job and contract search run as a loop rather than a panic: guided intake, honest 0–100 scoring, live-link verification, and a five-lens adversarial resume review per role. No server, no signup, and it never auto-applies.

Product · live

ROFLcopter live

Full-stack build · real-time multiplayer

A real-time multiplayer AI party word game: synchronized rooms, live scoring, and model-generated play running against latency budgets a turn-based system never has to meet.

Product · beta

4Vida beta

Program Module on GWEDU · bilingual (EN/ES)

An AI lifestyle and fitness coach shipped as a Program Module on GWEDU: its own repository of content and configuration, zero engine code. It is the proof the module model works: a completely different subject domain, brand, and persona set running on the same engine as a training course, with no fork.

Problems worth showing

The parts a demo does not reveal. Each of these is a decision that survived contact with production.

Spend that cannot run away

Public AI features fail on cost before they fail on quality. Every cloud call passes a budget gate with call caps and hard dollar ceilings, and visitors choose the question, never the provider or the spend.

The stronger guarantee is the ledger: a completion that cannot report readable token usage trips a persistent breaker that takes cloud routing offline until an operator clears it. That breaker has fired in production against a perfectly healthy response with a changed usage format, which is the design working. Unmetered spend is an outage, not a rounding error.

Status that tells the truth

The public status indicator is computed from whether a route can actually serve the next turn, not from whether keys happen to be configured. A system that says "online" while every route is dead is worse than one that admits it is degraded.

Ranking that has to prove itself

Retrieval quality is where hand-tuning goes to hide. The search behind the terminal is measured against a 73-question golden set built from real questions, each paired with the specific production chunk that should answer it, run over the live 63,105-chunk corpus. Adding a cross-encoder reranker moved recall@3 from 0.42 to 0.51, and that is the number that justified shipping it.

The same harness deleted work. Two hand-written query classifiers that nudged scores by substring looked reasonable and measured as noise: one fired on 0 of 73 questions, the other on 7, six of them false positives. They were removed rather than defended. The golden set and the eval script now live in the repository, next to the code they measure, and the harness fails loudly if it ever drifts from the search path it claims to be scoring.

Branding as configuration, not forks

Palette, logo, identity, intake questions, personas, routing policy, and copy load at runtime from a deployment bundle. One codebase serves a consulting terminal and a fitness coach without a branch between them, which is what makes the framework worth releasing publicly at all.

Research that publishes itself

A scheduled discovery pipeline polls feeds, repositories, topics, forums, and video channels; a scout scores what it finds; notable items flow to a public ticker, a curated digest, and the CODEX library. Featuring and publishing stay explicit operator actions with audit records. The automation gathers, a human decides.

How the work is done

Engineering practice, stated plainly, because it is the part that survives the handoff.

Discipline

Tested and reviewed

4,648 backend tests run on every pull request alongside frontend lint and build and an automated code review gate. A merge to master deploys production on its own, but only once CI on that exact commit is green; the manual path still exists for reruns and emergencies and still requires a typed confirmation.

Discipline

Secrets stay server-side

Model access runs through an on-host key broker with short-lived virtual keys, so rotating a provider credential is an operations task rather than a code deploy. Credentials never enter application code, logs, or model context.

Discipline

Operable by design

Scheduled jobs, backups, restore verification, and retention run on the box with their own logs and alerting paths; admin actions are token-gated and audited. The unglamorous half of shipping is treated as part of the build.

Credentials

Verifiable, and linked to their issuers.

Verified

Claude Partner Badge: Claude Code

Issued by Anthropic, 2026-07-09.

Certified

Claude Certified Architect: Professional

CCAR-P, issued by Anthropic, 2026-07-17.

Certified

Claude Certified Architect: Foundations

CCAR-F, issued by Anthropic, 2026-07-01.

Certified

Claude Certified Developer: Foundations

CCDV-F, issued by Anthropic, 2026-08-02.

All four are on one Credly profile, and the full history is on the CV.

Client engagements are covered by confidentiality. Detail on scope, structure, and references is available on request under NDA.