Ahmed Eltigani

Contract

• Sole developer of Express Ops, an operations platform for a tour operator: schema, role-based UI, OCR pipeline, CI/CD and production deployment.
• Used daily by the operations team; I estimate it saves them roughly 3–5 hours a day.

Employee

• UI/UX design and front-end development at a company building web3 technologies and trading bots.

• Sole developer of an operations platform used daily by a tour operator's team; I estimate it saves roughly 3–5 hours a day.
• Built a manifest pipeline (Mistral OCR) that reconciles extracted rows against printed footer totals and routes mismatches to an editable review queue; exports styled XLSX driver lists and DOCX customer lists; detects double bookings.
• Own CI/CD and security: dependency-audit gate, lint, tests, typecheck and build before an SSH deploy to Docker Compose; non-root multi-stage image, role-gated navigation and MFA.

Stack: Next.js, TypeScript, Convex, Mistral OCR, Docker Compose, GitHub Actions, Vitest

• Builder of an early-MVP multi-tenant SaaS for tour agencies: manifest ingestion with four parse providers and LLM extraction routed across five providers.
• Route generation on Google Maps or MapLibre with OSRM; driver dispatch over WhatsApp, Telegram and Slack with inbound status replies; Python FastAPI parsing service.
• Tenant isolation with Clerk and Convex ownership guards; Sentry across browser, server, edge and Python runtimes; separate local, staging and production stacks.

Stack: Next.js, TypeScript, Convex, Clerk, Python / FastAPI, Docling, MapLibre, Sentry, Docker

• Author of a five-crate Rust workspace that mediates AI agent tool calls: scope validation, operator approval via signed tokens, emergency stop and an audit log.
• Built the MCP client and interceptor layer (stdio and HTTP transports) and a ratatui terminal UI with session resume.
• CI runs tests, clippy and rustfmt; cargo-dist produces release builds for macOS, Linux and Windows.

Stack: Rust, MCP, Terminal UI, Cargo workspace

• Author of an async Python downloader for seven file hosts with a Textual TUI, a scripting CLI and an MCP server; published on PyPI (getit-cli) with a Homebrew tap.
• Concurrent downloads with resume and speed limits, MD5/SHA256 verification and client-side Mega.nz AES-CTR decryption.

Stack: Python, asyncio, Textual, SQLite, MCP

• Builder of a gamified, AI-coached exam-preparation app for medical graduates (Next.js, Convex, Clerk, TypeScript).
• Weakness analysis combining accuracy, confidence and recency drives practice sessions; includes the Clinical Poker pattern-recognition mini-game.

Stack: Next.js, Convex, Clerk, TypeScript

• Mobile-first Next.js page that sends customers from a QR code to Google, GetYourGuide, Viator or Booking.com in one tap.
• Per-channel click events for conversion visibility; channels configured in one file; built for one-click Vercel deployment.

Stack: Next.js, Tailwind CSS, Vercel Analytics

• Native Swift macOS menu-bar app that detects active calls and manages audio devices.
• Distributed with a Makefile and a Homebrew cask.

Stack: Swift, macOS, Xcode

• Co-developer on a team graduation project, working on the React front end and the AI integration (Gemini API and OpenAI Realtime API).
• Reading screen with in-context AI explanations and simplifications by text or voice, and a memory screen for saved quotes.

Stack: React, Tailwind CSS, Gemini API, OpenAI Realtime API

Nişantaşı University