Portrait of Amr
📍 Cairo, EgyptSelf-taught engineerAI-first
// 00WHO I AM

About Me.

I'm a self-taught software engineer based in Cairo, Egypt. No formal CS degree — I'm actually studying Economics and Political Science (GPA 3.3). The programming side I built entirely on my own.

That mix shapes how I think. Studying economics and political science taught me to see systems and incentives; hands-on experience in trade and marketing through family business taught me how real markets and real customers actually behave. I'm also genuinely fascinated by neuroscience — understanding how the human mind works is, to me, the closest thing we have to a blueprint for building real intelligence.

Teaching myself to code taught me to learn fast, debug from first principles, and ship without a safety net. I build AI-first — using AI tools as part of my workflow, the same way any senior engineer today uses their best tools. But every system I ship, I understand end to end, from database schema to the pixel the user actually touches.

I don't just want to build systems that work — I want to build ones that make things better for the people who use them. That's the kind of engineer I'm trying to be.

35+
Database Tables
MaxiCLI multi-tenant schema
220+
Test Files
Domain, schema & agent contracts
4
Architecture Layers
Domain / service / repo / API
// 02DEPLOYED · BUILT END TO END

Core Systems — Selected Work

Production-minded systems, one philosophy: async-first cores, ruthless isolation, and interfaces sharp enough to sell themselves.

SYS.01 — FINTECH

Trading Bot

Algorithmic execution · Realtime signals

Composable strategies feed a risk engine that wraps every signal; a backtest lab mirrors the live runtime exactly.

PythonpandasccxtTimescaleDB
12 strategies·<40ms exec
LiveSYS.02 — E-COMMERCE

Casa Cava

E-commerce · Wholesale & retail

A dual funnel: retail self-serves through the cart, wholesale converts to WhatsApp. An AI assistant answers product questions and quotes prices.

AstroTailwindCSSAI ChatWhatsApp API
8 categories·24/7 AI support
SYS.03 — AI INFRASTRUCTURE

AI Agents

Composable task-agent framework

Separates proposal from execution. Pydantic v2 tool contracts, policy guardrails, idempotent retries, and a replayable audit trail.

LLM OrchestrationTool-callingPydantic v2Streaming
30+ tools·0 unreviewed exec
// 03CAPABILITIES

The Stack I Ship With, by Layer

A working stack rather than a wishlist — grouped by layer, no invented proficiency scores.

AI / LLM

LLM OrchestrationRAGpgvector
  • Tool-calling
  • Pydantic v2
  • Streaming
  • Agent runtimes

Backend

PythonFastAPIPostgreSQL
  • Redis
  • TimescaleDB
  • Alembic
  • Async I/O
  • Docker

Frontend

Next.jsAstroReact
  • TypeScript
  • Tailwind
  • Three.js
  • WebSockets

Infra & Ops

VercelGitHub ActionsLinux
  • Nginx
  • Stripe
  • Supabase
// 04SECURE CHANNEL — OPEN

Bring a hard product problem. Let's build something.

AI systems, automation, SaaS, and product engineering — where the result has to be both useful and memorable.

OPEN TO NEW PROJECTS · 2026

Every engagement starts with a technical deep-dive: your constraints, your concurrency profile, your data model. Then we build — fast, measured, production-grade.

CAIRO → GLOBAL · UTC+2 · RESPONSE < 48H
amr@architect:~/contact — secure-tty

$ ./initiate_contact --priority high --channel direct