Mohamed Achkari BegdouriSoftware Engineer · Morocco

I follow difficult problems through every layer.

Mohamed Achkari Begdouri

Across real-business work, SaaS products and technical research, I move from interface and services to data, infrastructure, AI and networks—whatever the problem needs.

InputA difficult problem
Interface

Make the problem understandable and the next action obvious.

OutputA working system
Selected work / 01—05

The work is the evidence.

Five projects. Five different systems. One recurring habit: understand the problem before choosing the technology.

01Real-business partnership

Golden Sweet

A website became ongoing technical ownership.

A multilingual product-discovery experience for a healthy pastry business in Tétouan, connected to search, Maps and WhatsApp—and evolving toward AI-assisted dietary discovery.

01

Discover

Dietary categories, product information, Arabic + English

02

Reach

SEO, Google Maps and WhatsApp touchpoints

03

Operate

Hosting, monitoring, maintenance and ownership

04

Evolve

Direction toward AI-assisted product discovery

Evolution, not a finish line

The AI assistant is presented as a next direction for navigating structured product information—not as deployed medical guidance.

02Connected production system

coWorkflow

The interface is only the visible edge.

01

Typed interface

React · Next.js · TypeScript

02

Validation + identity

Zod · React Hook Form · Passport.js

03

Application services

API routes · TanStack Query

04

Relational data

PostgreSQL · Drizzle ORM · Supabase

05

Storage + schedules

Cloudflare R2 · Cron jobs

06

Production

Visualization · Vercel deployment

One product, connected across every production layer. No invented dashboard statistics—just the architecture that makes the interface possible.

03Decentralized VPN research

How much of a route should any one participant know?

An investigation into progressive multi-hop routing, distributed trust, node selection and the privacy-performance trade-offs of a decentralized network.

Interactive research model

Construct the route one decision at a time.

  1. Client: requests a target hop count
  2. Entry: selects a suitable next node
  3. Relay: continues the route progressively
  4. Exit: connects to the public internet
  5. Destination: sees the exit connection

Step 1 of 5. The client starts with a desired route depth.

This visualization explains an explored architecture. It does not claim that anonymity, traffic-analysis resistance or exit-node risk has been solved.

Trust boundary

A participant should not need the complete route to forward traffic.

Selection tension

Latency, bandwidth, reliability and reputation compete with privacy goals.

Open risk

Exit responsibility, abuse and traffic fingerprinting remain hard questions.

04Code Execution Platform · Containerized runtime

Run is a lifecycle, not a button.

A Next.js and Spring Boot programming environment that routes Python, Java, C and C++ submissions through temporary Docker execution environments.

Execution lifecycle01 / 08
request accepted
submission.pycontainer_tmpresult
05Falcon Wings · Custom-printing business · Tangier

A digital front door for physical production.

Website improvements for a custom-printing business in Tangier, translating services, work examples and production steps into a clearer customer journey.

  1. 01Browse the work

    Gallery and service presentation

  2. 02Understand the process

    Structure and production explanation

  3. 03Start a conversation

    Responsive WhatsApp contact path

Operating range

The stack follows the problem.

Interface, backend, data, infrastructure, networks and AI appear as parts of systems—not as a wall of technology logos.

01

Interface

React · Next.js · TypeScript

Golden Sweet / coWorkflow
02

Services

Node.js · Express · Spring Boot

coWorkflow / Code Execution
03

Data

PostgreSQL · MongoDB · Supabase

Application systems
04

Infrastructure

Docker · Vercel · Cloudflare R2 · AWS

Operate + deploy
05

Networks

Routing · virtual networks · trust

DVPN research
06

Intelligence

AI assistants · semantic discovery · ML

Useful integration
Beyond implementation

Quality is part of the architecture.

Software quality

Mohamed is pursuing a Master’s degree in Software Quality at the Faculty of Sciences of Tétouan, Université Abdelmalek Essaâdi, extending practical work with maintainability, validation, reliability and structured engineering.

Ownership beyond code

During the 2025–2026 academic year, he served as President of the InfoM4th Computer Science Club, coordinating teams, university stakeholders, speakers, sponsors and technical programs.

Cloud foundation

AWS Academy Cloud Foundations complements hands-on deployment work with IAM, EC2, VPC, cloud networking, security and architecture concepts.

Next system

Bring the problem that crosses layers.

Open to serious software-engineering roles and thoughtful technical collaborations.

Contact details pending.No email address or professional profile was supplied, so no channel has been invented.

Review selected work