I build production systems that solve expensive problems. Not demos. Live platforms — tested, deployed, and running. Each repo below is a working system someone can put to work today.

The work ships under the organvm ecosystem I architect — 238 public repositories (209 original), 29,811 contributions in the last year. On GitHub since 2016.

Each is a working production system — source, tests, and live deploy. The numbers are the repository's own CI output, not adjectives.

A browser userscript that exports your AI chat conversations into five portable, archival-quality formats — Markdown, HTML, JSON, PNG, and plain text — entirely in-browser. No account, no server, no data leaves your machine; it reverse-engineers the platform's internal API and linearizes the branching message tree, with a provider layer live for ChatGPT and scaffolded for Claude and Gemini.

Solves — ChatGPT conversations are ephemeral by design. There's no native export to Markdown or HTML and no bulk download — the official dump is one unwieldy JSON blob most people can't use. Knowledge you paid for in thinking time evaporates or gets locked in a format you can't pipe anywhere.

170 tests across 16 suites · ~5,900 lines of TypeScript · distributed on GreasyFork, GitHub, Docker (GHCR), and Cloudflare Pages · 5 export formats · 9-language UI · MIT-licensed · 549 commits, actively maintained

→ Build this into my product  ·  Work with the builder

A production platform that collects UCC-1 filings from all 50 Secretary-of-State portals, enriches each record with SEC / OSHA / USPTO / Census and key-gated premium sources, and scores every prospect 0–100 on financing likelihood with an A–F health grade — delivered via dashboard, REST API, and CLI.

Solves — Commodity aged UCC lists are cheap and worthless. Exclusive, fresh, enriched, scored UCC leads are the difference between a dialer full of dead numbers and a funded deal.

3,399 passing tests · Terraform AWS (multi-AZ RDS, Redis, S3) · 60+ collection agents · all 50 states · live deploy

→ Deploy this for your shop  ·  Work with the team that built this

A production-grade enforcement layer for commitments with real financial stakes: Stripe FBO escrow, a double-entry ledger, peer-audited proof review, bounty incentives, honeypot QA, and privacy-firewalled B2B funding where employers can sponsor engagement without seeing individual health data.

Solves — Accountability and corporate-wellness programs spend money on intent but do not create follow-through. The proof is usually self-reported, the incentives are weak, and holding personal health data creates liability.

1,107 tests across workspaces · public pitch surface returns 200 OK · NestJS + Next.js + React Native + Tauri monorepo · Stripe escrow and double-entry ledger · CodeQL, secret scan, Terraform, beta-readiness, and e2e CI gates

→ Deploy this for a commitment program  ·  Work with the team that built this

A 'phygital' beauty platform: hyper-realistic AR/3D virtual try-on, medical-grade skin diagnostics, AI style coaching, and salon booking in one app. Two modes — clients try on looks and book; stylists run a dashboard, manage consultations and proposals, and coordinate their team.

Solves — A salon visit costs hours and real money, and a wrong cut or color carries a six-to-eight-week regret cycle — so customers hesitate and bookings stall. A brand with no try-on misses the digital-first shopper entirely.

~42,800 lines of TypeScript · 158 test files · 60 feature components across a 40-tab app · Stripe + Lemon Squeezy billing · live on Netlify · Docker + Vercel + Netlify + Pages deploy

→ Build try-on into my product  ·  Work with the builder

A production system that sweeps Gmail and Apple Mail across accounts, classifies every thread against a protocol ladder, tracks every owed reply in an obligations ledger, and composes voice-matched draft replies — draft-only, it never sends. Deploys as a Cloudflare Worker; runs keyless (IMAP app-password / Apple Mail osascript).

Solves — Obligations fall through the cracks — the escalation you missed, the reply you owed a week ago — while triage burns the most expensive hours of the day. Every 'AI email tool' wants send authority nobody sane will give it.

400+ tests · Cloudflare Worker deploy (wrangler) · Gmail + Apple Mail, keyless · draft-only by design — composes replies, never sends · obligations ledger — every owed reply tracked

→ Run this on my inbox  ·  Work with the team that built this

A living ladder of self-* organs (self-sustaining, self-routing, self-feeding, self-healing) that run autonomously on a heartbeat. The fleet maintains itself: monitors liveness, reaps stale work, syncs state, escalates failures. Built-in organs: proprioception, VIGILIA federation, branch-reap, clone-lifecycle, credential durability, health monitoring.

Solves — Every agent fleet needs: health checks, credential refresh, stale-work cleanup, state reconciliation, failure escalation. If manual, each system costs ~2h/week ops. If ignored, cascading failures ripple through the fleet.

9-rung organ ladder (each self-* owns its own domain) · Heartbeat-driven autonomic loop (scripts/metabolize.sh) · Self-healing on regression (heal/ branch pattern) · Live: proprioception organ, VIGILIA federation, branch-reap, credential-wall · Zero manual upkeep — runs forever once seeded

→ Adopt self-healing fleet ops  ·  Work with the team that built autonomic infrastructure

A public portfolio site showcasing 16 completed production systems — each with a case study, real architecture diagrams (p5.js interactive), live demos where applicable, and the decision trees behind the build. Built with Astro, deployed on Netlify, fully indexed and discoverable.

Solves — A portfolio of demos and one-liners doesn't prove judgment. Seeing the tradeoffs, the decisions, the outcomes, and the architecture is the difference between 'built something' and 'ships production systems'.

609 files · Astro + TypeScript · p5.js interactive visualizations · 69 test files across the portfolio + content · 16 case studies with architecture, decisions, and outcomes · Live demos & interactive diagrams · Netlify deployment + CI/CD · SEO-optimized content sites

→ Explore the case studies  ·  Study the builder's thinking

I harvested the profile READMEs of all 61 accounts I follow and rebuilt their best techniques as my own self-hosted, no-third-party-widget versions. The most-adopted signals across my follows:

• Shields.io badge row — 8 of my follows use it (@prometheus, @gollum, @asciinema) → my own self-hosted badges.svg.
• Language breakdown card — 2 of my follows use it (@madhanio, @JawherKl) → my own self-hosted languages.svg.
• GitHub stats card — 2 of my follows use it (@madhanio, @JawherKl) → my own self-hosted stats-card.svg.
• Contribution streak — 2 of my follows use it (@JawherKl, @mennylevinski) → my own self-hosted streak.svg.
• Activity line graph — 1 of my follows uses it (@JawherKl) → my own self-hosted heatmap.svg.
• Contribution snake animation — 1 of my follows uses it (@mennylevinski) → my own self-hosted snake.svg.

Full ranking: assets/follow-harvest-report.md · regenerated weekly.

Have a problem one of these solves? — Deploy it for your shop

Pick the depth that fits. We feed you the output, run it under your brand, build it for your exact world, or become your engine.

Hiring someone who ships at this level? — Work with the builder

Everything here is the portfolio. If you need a senior builder who owns systems end-to-end — data, infra, AI, deploy — this is the evidence.

If any of this fits, reach out. This conversation starts at serious.

_{Every number on this page is regenerated from the live GitHub API and re-derivable from stats-manifest.json (api-attested = re-run the query; repo-attested = the repo's own CI). Visuals are self-hosted SVGs rendered by organvm/limen — no third-party widgets. Generated 2026-07-18T15:07:27Z.}