Skip to content
N.
N. Slimani — Systems Engineer · Cloud Architect · ML Researcher
Open to collaboration • Research • Projects • Opportunities

Systems Engineer · Cloud Architect · ML Researcher

N. Slimani

I design and secure the infrastructure organisations depend on, and build machine learning systems that make sense of their data. 5+ years of enterprise infrastructure, cloud architecture, and applied AI research, based in Budapest, Hungary.

AWSAzureKubernetesDockerPythonJenkins
Scroll
145+Enterprise users in managed infrastructure
40%Operational efficiency improvement via automation
85%NLP classification accuracy on ML thesis
45%Customer engagement lift from cloud automation

// 01 — WHO I AM

Background & experience

I'm a Systems Engineer and Cloud Architect based in Budapest, Hungary. I build and secure the infrastructure that keeps organisations running — and what drives me is the intersection of reliable systems and intelligent automation.

My academic background in Computer Engineering (Master's, Excellent distinction) informs how I approach infrastructure: not just as plumbing to maintain, but as a platform to build intelligent, data-driven solutions on top of.

Reliability first

Systems should be boring in the best way — predictable, observable, and recoverable before they're clever.

Research-minded

I read papers, run experiments, and publish results. Curiosity is a working tool, not a hobby.

Automate the toil

If a task repeats, it becomes a script, then a pipeline. Manual process is a bug to be fixed.

Always learning

New certification, new framework, new dataset — the stack keeps moving and so do I.

Portrait illustration of Nasreddine Slimani

Currently

Administering enterprise infrastructure at Diamond Diagnostics Inc. while publishing federated-learning research on the side.

// 02 — TECH STACK

Skills & expertise

Competencies built across enterprise deployments, open-source projects, and academic research. Hover — or tap — a card for a quick note on how I use it.

Cloud & Infra

DevOps

Security

Programming

Networking

Data & Monitor

ML / NLP

Certifications

  • AWS Cloud Practitioner

    Cloud Architecture Fundamentals

    2023

  • DevOps Bootcamp

    Techworld with Nana, CI/CD

    2024

  • Linux Admin / DevOps

    Nix Tech Company

    2024

Languages

  • ArabicNative
  • EnglishProfessional
  • FrenchProfessional

// 03 — FEATURED WORK

Projects

Engineering projects spanning infrastructure automation, machine learning applications, and cloud architecture.

01 / Featured Build · 2024

Stock Management System

The problem

Multi-location inventory was tracked manually across spreadsheets — slow to update, error-prone, and impossible to audit quickly when stock moved between sites.

My solution

A full-stack inventory platform with a scalable MongoDB data model built for multi-location stock, paired with an automated document pipeline for invoices and purchase orders — all exported as browser-native PDFs with no third-party service in the loop.

PythonNodeJSMongoDBPDF Generation
  • Scalable MongoDB architecture handling multi-location inventory
  • Automated invoice and purchase order generation pipeline
  • Browser-native PDF export without third-party services

Designing the data model around how stock actually moves (not how it's reported) made every feature after it — audits, PDFs, multi-location views — fall out almost for free.

View project
02 / Research Build · 2023

NLP Sentiment Analysis

The problem

Understanding public sentiment at scale from raw, noisy social text requires a model that generalises past hand-tuned keyword rules.

My solution

A deep learning sentiment classifier trained on Twitter data, paired with a real-time interactive visualization dashboard so results are legible to non-technical stakeholders, not just the model.

PythonNLPDeep Learning
  • 85% classification accuracy
  • Real-time interactive visualization dashboard

The dashboard mattered as much as the model — a good classifier nobody can interrogate is a lot less useful than a decent one people can actually explore.

03 / Production Build · 2023

UrBoutik.com

The problem

A growing e-commerce operation needed production-grade infrastructure and marketing automation without enterprise budget or headcount.

My solution

A production e-commerce platform deployed on AWS with SES-powered marketing automation, high-availability infrastructure, and DNS optimisation tuned for real customer traffic.

WordPressAWS SESCloud Infra
  • 45% customer engagement improvement
  • High-availability cloud architecture

Most of the engagement lift came from infrastructure being invisible — fast pages and reliable email deliverability, not a redesign.

Daily tools I've built

// 04 — ACADEMIC RESEARCH

Research & experiments

Published and ongoing ML research with a focus on federated learning, privacy-preserving NLP, and low-resource language modelling.

AIDAM26 · 2026

Federated Fine-Tuning of DziriBERT

Privacy-preserving NLP for Algerian dialect sentiment analysis using federated learning strategies on a low-resource language model. Compares FedAvg, FedProx, semi-supervised, and distilled approaches under realistic non-IID client distributions.

NLPFederated LearningDziriBERTLow-resource AI
FedAvg reached 84.8% accuracy vs centralized baseline
FedProx improved convergence under non-IID client data
SSFL reduced labeling requirements while maintaining performance
DistilBERT offered a lighter alternative with competitive results
View paper & results
PAIS26-IEEE · 2026

Automated Privacy-Preserving FL for IoT Intrusion Detection

Interactive research presentation on a privacy-preserving federated learning pipeline for IoT intrusion detection. The work focuses on automated deployment, non-IID robustness, low-bandwidth communication, and near-centralized performance without exposing raw device data.

Federated LearningIoT SecurityPrivacy-preserving MLIntrusion Detection
Near-centralized detection performance with privacy-preserving training
Automated pipeline reduces manual FL setup and deployment effort
Very low communication overhead using parameter exchange only
View paper & results
ASREM26 · 2026

XAI-IDS: Explainable Intrusion Detection for Smart Energy IoT

Explainable intrusion detection framework combining Random Forest, XGBoost, and LightGBM with SHAP TreeExplainer to deliver global, local, and per-class explanations on the RT-IoT2022 benchmark, with a focus on energy IoT edge-gateway deployment.

Explainable AISHAPIoT SecurityXGBoost
All models reached macro F1 above 0.98 on RT-IoT2022
4-feature SHAP consensus shared across all three models
88% feature reduction with under 0.007 F1 loss for edge deployment
View paper & results

Next paper

Currently in progress. Check back soon.

In Progress

// 05 — OPEN SOURCE

Open source

Public repositories and daily tools, pulled live from GitHub.

@noahswitch

// 06 — EXPERIENCE

Career & education

Systems Engineer & Cloud Administrator

Diamond Diagnostics Inc. Sep 2022 – Present Budapest, Hungary

Primary administrator for a 145+ user enterprise environment, managing Windows AD lifecycle, Office 365, cloud architecture, cybersecurity operations, and ITIL-compliant change management.

Cybersecurity Intern

DEKRA Magyarország KFT. Mar – Aug 2022 Budapest (Hybrid)

Vulnerability assessments, security procedure documentation, and policy review support under senior cybersecurity staff.

Self-Employed IT Consultant

N Slimani EV Sep 2021 – Sep 2022 Remote

Delivered software development and sysadmin services to 5+ clients. Built e-commerce solutions with 40% operational efficiency gains and automated inventory systems.

Education

M.Sc. Computer Engineering

Széchenyi István University, Győr, Hungary

2020 – 2022Excellent

Thesis: "Sentiment Analysis of Twitter Data Using Neural Network Models"

#hpc#cloud-architecture#network-security#ai#distributed-systems

B.Sc. Computer Systems

University of Bechar, Algeria

2017 – 2020Good
#database-systems#networks#os#algorithms

Stipendium Hungaricum Scholarship

Hungarian Government, 2020 — Full scholarship for Master's studies

// 07 — REACH ME

Let's build something
meaningful together.

Whether it's a senior role, a research collaboration, or just a conversation — I'm always happy to connect with people building interesting things.

Budapest, Hungary
Full-time opportunitiesResearch collaborationsFreelance projectsOpen-sourceStartup ideas
Email Me

Also find me on

LinkedInGitHubResearchGateTelegramBlog