Skip to content
Tarek Okasha

About

Tarek Okasha.

I'm a software engineer in Cairo. I build AI systems, automations, and custom internal software for founders and operating teams — the tools a business runs on every day. I take a handful of projects a year and build each one properly: tested, documented, and handed over with an architecture the next engineer can actually read.

I'm not the right call for a vague idea you want someone to figure out. I'm the right call when the problem is real and expensive — a process held together by spreadsheets and goodwill, a workflow quietly eating your team's week, a system your business needs that no vendor sells. Bring me that, and I'll build the thing that makes it go away.

Background

I studied Robotics and Automation Engineering at the German International University. The degree gave me the foundations — control systems, embedded hardware, sensor fusion. The real education came from the things I built while still a student: starting with an auction-pricing model for a friend who resold electronics, a system that taught me more about production constraints in one month than a semester of coursework. Most projects since have had the same shape: a real constraint, limited resources, and a hard requirement that it actually works.

Alongside the robotics work, I was shipping paid software projects — the first at 20. I went independent before graduating because the client work was already more demanding and more interesting than any job I could see ahead of me. Seeing something you built running in a real business, with real people depending on it, is a kind of feedback that sharpens your judgment faster than anything else.

I'm 22. I work with clients across the Gulf, Egypt, and globally. Almost all of them came through referral — which, in my experience, is the only honest measure of whether the work was actually good.

Tools

AI & LLMs

  • Claude (Anthropic)
  • GPT-4o / o1
  • Gemini
  • Fine-tuning
  • RAG pipelines
  • Vector search
  • Prompt engineering
  • Evals & benchmarking

AI Engineering

  • LangChain
  • LlamaIndex
  • CrewAI
  • Computer Vision
  • Embeddings
  • Function calling
  • Pinecone / Weaviate
  • OpenAI Assistants API

Automation & Integration

  • n8n (self-hosted)
  • Zapier
  • Make (Integromat)
  • GoHighLevel
  • WhatsApp Business API
  • HubSpot API
  • Salesforce API
  • Webhook orchestration

Frontend & UI

  • Next.js 15
  • React 19
  • TypeScript (strict)
  • Tailwind CSS v4
  • Framer Motion
  • GSAP
  • Three.js
  • Figma → code

Backend & Data

  • Node.js
  • Python / FastAPI
  • PostgreSQL
  • Redis
  • Supabase
  • Prisma ORM
  • REST & GraphQL
  • Stripe

Infrastructure

  • Vercel / Netlify
  • Docker
  • Linux (Ubuntu)
  • GitHub Actions
  • AWS (EC2, S3, Lambda)
  • Cloudflare
  • CI/CD pipelines
  • Monitoring & alerting

How I work

Four things I'd put in writing.

01
01

I write the code.

No subcontractors, no 'team of vetted freelancers.' When you hire me, I'm the person on the keyboard. You get my thinking on your problem, not a project manager who escalates to someone you've never met.

02
02

Working software in weeks, not pitch decks in months.

I ship weekly demos. You see progress. No surprise invoices at the end of month three. If something isn't working, we know in week two — not week twelve.

03
03

If it doesn't move a number, it doesn't ship.

Every system I build has a metric attached. Saved hours. Reduced errors. Faster cycle time. If we can't agree on what success looks like before we start, we don't start.

04
04

I say no a lot.

I take on 6 projects a year. Not 20. I'd rather recommend someone better suited than over-commit and underdeliver. Scarcity isn't marketing — it's how good work actually gets done.

Outside work

I lift heavy things and box at the gym — the most consistent thing I do. I read more nonfiction than is probably necessary — currently working through The Systems Mindset: Managing the Machinery of Your Life, which keeps turning up in my work in ways I didn't expect. I'm slowly learning to make a proper espresso. The machine judges me.