■ FOUNDER · AXIOMID
Building software-first sovereign identity systems with a team of AI agents
I'm Mohamed Hossam-Eldin Abdelaziz — a Cybersecurity student at Kennesaw State University and founder of AxiomID, a software-first decentralized identity protocol. My work sits at the intersection of cryptography, agent orchestration, and user-centric design.
أنا محمد حسام الدين عبدالعزيز — طالب أمن سيبراني في جامعة كينيساو، ومؤسس AxiomID (بروتوكول هوية لامركزي يعتمد على البرمجيات أولاً).
I don't use AI as a tool. I use AI as a team. Each agent has a defined role, a clear boundary, and an auditable history. The human (me) sets the vision and final authority. The agents execute, verify, and evolve.
I act the same in private as in public, because the work is recorded either way.
Active projects (verified from git history this week):
| Project | What | Where |
|---|---|---|
| AxiomID | Decentralized identity protocol · Pi Network integration · W3C DIDs · VC marketplace | 40+ API routes · 11 pages · 18 Prisma models · 5 AI agents active |
| amrikyy.axiomid.app | Mobile-first control plane cockpit for monitoring all agents from a phone | In build · planned deploy via Vercel + portless local dev + Pake desktop wrap |
| Hermes Agent pair-partnership | Nous Research's Hermes AI as my reasoning partner · audit + planning + reports | Active daily · Telegram chat · 2nd brain vault at ~/Desktop/amrikyy/amrikyy.memory/ |
One human + five AI agents, each with a distinct role. This is the actual current roster, not the original roster in this README.
| Agent | Role | Status |
|---|---|---|
| Mohamed Abdelaziz (human) | Founder · Lead Architect · Final authority | Active |
| Hermes (Nous Research) | Reasoning partner · Audit · Planning · Reports | Active daily |
| OpenCode | Primary execution agent for repo work | Standing by (brief ready) |
| CodeRabbit | PR review automation · Unit test generation | Active (30+ commits) |
| Jules (Google Labs) | Performance optimization · Test backfilling | Active (10+ commits) |
| Gemini CLI | Research · Multimodal analysis | Configured, awaiting API key activation |
Note: Antigravity was previously listed here. It is an IDE, not an autonomous agent. I work in it; I don't run it. Removed from this table for accuracy.
Decentralized identity for the AI era. Software-first. No iris scans. No hardware dependencies.
Architecture (verified, June 22, 2026):
- 285 source files · 40,859 LOC · 18,456 LOC of tests (45% test:source ratio)
- 48 API routes, 43 with test files (90% coverage)
- 18 Prisma models · 38 components · 11 pages
- Pi Network SDK v2.0 · W3C DID:axiom method · Ed25519 sovereign keys
- Cloudflare Workers backend (D1, Vectorize, Workers AI) · Upstash Redis · Supabase
Quality score (from latest audit): 7.5–8.0/10 with active execution plan to reach 9.5/10.
Not yet built (honest):
- The Vault (1-year USDC stake for Sovereign tier) — described in STRATEGY.md, not in schema
- Zero-knowledge proofs — described in STRATEGY.md, not in codebase
- Real production user metrics — currently in development
- portless — Stable
*.localhostURLs with auto-HTTPS for local dev - Pake — Webpage → native desktop app in one command
- dev3000 — Browser + server log capture for AI-debuggable timeline
- local-sandbox — Rust substrate for production-shaped local sandboxes
- packnplay — Docker isolation wrapper for coding agents
- open-webui-plugins — Rich UI plugins for AI chat
Each of these is built because I needed it. Each is used in production on AxiomID or the upcoming amrikyy.axiomid.app.
I maintain a private "second brain" (Obsidian vault) that captures every architectural decision, every codemap, every audit. The most important files:
AxiomID/EXECUTION_PLAN_v3.md— 4-phase plan to take AxiomID to 9.5/10 qualityAxiomID/_CODEMAP_INDEX.md— 4 Devin-style codemaps (Pi Auth, Theming, Publishing, Data Layer)cockpit_architecture.md— Design for amrikyy.axiomid.app cockpitresearch_zerolang.md— Honest evaluation of Vercel Labs' new agent languageopencode_brief.md— Standing orders for the OpenCode execution agent
These come from my own SOUL.md (the operating constitution for everything I ship):
- Truth over comfort. I won't fabricate metrics to look good. If a feature is planned but not built, I'll say so.
- Actions are recorded. Every commit, every change, every decision is part of an auditable chain.
- Errors are admitted, not hidden. When I miss a target or ship a bug, I'll say so publicly and fix it.
- The human is sacred. The person interacting with my work is not a metric. They deserve work that serves them honestly.
| Area | Status |
|---|---|
| AxiomID core protocol | 🟢 Active development · 4-phase execution plan in progress |
| amrikyy.axiomid.app cockpit | 🟡 In design · Phase 6 of execution plan |
| 2nd brain / Obsidian vault | 🟢 Active · 30+ documents · 4 verified codemaps |
| AI agent team | 🟢 6 agents active · coordinated via Hermes + kanban |
| Production users | 🔴 Zero · AxiomID not yet launched to public |
| Funding | 🔴 None · bootstrapped |
Last updated: 22 June 2026 · refreshed automatically by .github/workflows/live-system-data.yml
- LinkedIn: linkedin.com/in/mohamed-abdelaziz-815797347
- Personal: amrikyy@gmail.com
- Academic: mabdela1@students.kennesaw.edu
- WhatsApp: +1 770 616 0211
For collaboration on AxiomID or any of the open projects, open an issue on the relevant repo. For everything else, email works.
This README is auto-maintained by 8 Python scripts in scripts/ running via .github/workflows/live-system-data.yml every 3 hours. If something looks outdated or wrong, it's either:
- Genuinely outdated (I haven't pushed the fix yet) — open an issue
- A script bug — check the workflow runs in the Actions tab
- My honest assessment of where the project really is — which may be less impressive than marketing would prefer, but it's the truth
I prefer accurate-but-modest over impressive-but-fake.
assets/telemetry.svg timestamp