Aymen Rahali

Aymen Rahali, AI & DevOps Engineer at VegaNext

I build AI products and keep them running.

I integrate LLMs, RAG chatbots and voice AI into real products, then deploy them with Docker, Nginx and CI/CD and monitor them in production.

View projects Download CV Contact

Open to AI Engineer, MLOps and DevOps roles, remote or with relocation

One chatbot request, end to end
  1. NginxTLS, reverse proxy
    run
  2. API serviceNode.js in Docker Compose
    build
  3. Retrieve contextvector store, RAG
    build
  4. Generate answerLLM API
    build
  5. Metrics and logsproduction monitoring
    run

Every step of this request is something I either build or run, shipped by CI/CD to Linux servers.

Measured results

  • −70% manual testing time Tessi
  • 3× deployment frequency Tessi
  • −40% build time Ipact Consult
  • −25% production bugs Ipact Consult
  • −30% project delays Sopra HR Software
  • 12 applications handled per month 3S

Across 5 professional experiences and backed by 3 NVIDIA certifications.

I like owning a feature end to end

I'm an AI & DevOps engineer who likes owning a feature end to end, from the first line of code to the moment it runs reliably in production.

At VegaNext (California, USA), I work as a Full-Stack AI Engineer on B2B SaaS products. I build features across React/TypeScript front-ends and Node.js/Python back-ends, and integrate AI into them:

  • LLM APIs in production features
  • a RAG chatbot that answers from company knowledge
  • voice AI for VegaTel, our cloud VoIP platform
  • automated workflows with N8N and AI agents

I also deploy and run what I build: Linux servers, Docker Compose, Nginx reverse proxy with SSL, CI/CD pipelines, and production monitoring.

Before VegaNext, I completed four engineering internships:

  • At Tessi, I built a test automation platform that cut manual testing time by 70%.
  • At Ipact Consult, I set up a full DevOps pipeline that cut build time by 40%.
  • At Sopra HR Software, I built an ML task-assignment system that cut project delays by 30%.
  • At 3S, I built a recruitment platform handling 12 applications a month.

I'm NVIDIA-certified in Deep Learning, LLM customization and Transformer-based NLP, and I'm looking for ambitious teams where AI creates real, measurable value.

Experience

  1. Nov 2025 – now

    VegaNext

    Full-Stack AI Engineer

    California, USA. B2B SaaS.

    I develop product features end to end, integrate AI into them, and deploy and monitor them in production.

    • Build and ship product features across React/TypeScript front-ends and Node.js/Python back-ends.
    • Built VegaNext's products and sites: VegaTel, Vega360, VegaMSP, VegaNext.com and MyLaffApp.
    • Integrate LLM APIs into production features.
    • Built a RAG chatbot that answers user questions from company knowledge.
    • Develop voice AI features for VegaTel, the company's cloud VoIP platform.
    • Automate business workflows with N8N and AI agents.
    • Deploy and operate services on Linux servers with Docker Compose, Nginx reverse proxy and SSL.
    • Set up CI/CD pipelines for automated deployments, and monitor production apps.
    VegaTel product target: 99.9% uptime

    Stack: React, TypeScript, Node.js, Express, Python, LLM APIs, RAG, N8N, Docker Compose, Nginx, CI/CD, Linux

  2. Feb – Sep 2025

    Tessi / Owliance

    Full-Stack & AI Engineer, final-year internship (PFE)

    Tunis. Project: DataFlow-test.

    • Built a non-regression test automation platform integrated with Jira Xray, cutting manual testing time by 70%.
    • Designed a microservices architecture with a Docker-based CI/CD pipeline, tripling deployment frequency.
    • Trained an NLP model that prioritizes critical test cases, focusing QA effort on high-risk areas.
    −70% manual testing time3× deployment frequency

    Stack: React, Spring Boot, FastAPI, Flask, MongoDB, Docker, NLP/Python, N8N

  3. Jun – Sep 2024

    Ipact Consult

    DevOps / Full-Stack Intern

    Tunis

    • Built a complete DevOps pipeline: Jenkins CI/CD, SonarQube quality gates, Nexus artifact management, Grafana and Prometheus monitoring.
    • Reduced build time by 40% and production bugs by 25%.
    −40% build time−25% production bugs

    Stack: Spring Boot, Docker, Jenkins, SonarQube, Nexus, Grafana, Prometheus

  4. Jul – Aug 2024

    Sopra HR Software

    Full-Stack & AI Intern

    Tunis. Project: SmartAssign.

    • Built SmartAssign, a task management system that auto-assigns work with machine learning based on skills, availability and priority.
    • Reduced project delays by 30%.
    −30% project delays

    Stack: React, FastAPI, Flask, Python, Machine Learning

  5. Jun – Sep 2023

    3S Standard Sharing Software

    Web Developer Intern

    Tunis

    • Built a job and internship platform handling 12 applications per month.
    • Added automatic recruit assignment, improving onboarding by 15%.
    12 applications a month+15% onboarding

    Stack: ASP.NET MVC, C#, JavaScript, Symfony

Selected work

AI features I built and the systems that run them.

AIProductionDevOps

VegaTel AI and VegaNext AI features

Company work at VegaNext, 2025 to now

AI features for a cloud enterprise VoIP platform and B2B SaaS products, built and deployed end to end.

  1. Problem

    VegaNext builds B2B SaaS products and VegaTel, a cloud enterprise VoIP platform. The goal: add AI features (LLM-backed features, a chatbot that answers from company knowledge, voice AI) and run them reliably in production.

  2. Architecture

    Client traffic goes through Nginx with SSL to Dockerized services, which call an LLM API and a vector store. Every service is monitored.

  3. My role

    • LLM API integration into product features
    • RAG chatbot over company knowledge
    • Voice AI features for VegaTel
    • N8N and AI-agent workflow automations
    • Docker Compose deployment behind Nginx with SSL
    • CI/CD pipelines and production monitoring

  4. Results

    Features shipped to production on VegaNext products and VegaTel, deployed and monitored by me. VegaTel targets 99.9% uptime as a product.

Visit vegatel.us

Architecture

High-level view. Internal code and data are not shown.

AIDevOpsMicroservices

DataFlow-test

AI-assisted test automation at Tessi, 2025

Non-regression test automation platform with Jira Xray integration and an NLP model that prioritizes risky test cases.

−70% manual testing time3× deployment frequency

Stack: React, Spring Boot, FastAPI, Flask, MongoDB, Docker, NLP, N8N

Read the case study

  1. The pain

    Manual regression testing was slow and spread QA effort evenly, whatever the risk.

  2. The solution

    • Microservices architecture (React, Spring Boot, FastAPI, Flask, MongoDB)
    • Docker-based CI/CD pipeline
    • Non-regression test runs integrated with Jira Xray
    • NLP model that ranks critical test cases first

  3. Results

    Manual testing time cut by 70%, deployment frequency tripled, QA effort focused on high-risk areas.

AI / ML

SmartAssign

ML task assignment at Sopra HR Software, 2024

Task management system that auto-assigns work using ML on skills, availability and workload.

−30% project delays

Stack: React, FastAPI, Flask, Python, ML

Source code on GitHub
Read the case study

  1. Inputs

    Each person's skills, availability and current workload.

  2. Machine learning

    Matches each task to the best-placed person.

  3. Assignment and results

    Tasks are assigned automatically. Project delays dropped by 30%.

Also built at VegaNext

Products in production, alongside VegaTel

  • Vega360

    Web app

    One portal for VegaNext clients, partners and staff: client projects, tickets, subscriptions and invoices, a partner service catalogue with orders, and internal project management (PSA).

    vega360.com

  • VegaMSP

    Product site

    Site for VegaNext's managed IT and security offer: managed network, endpoint security, VoIP and unlimited helpdesk.

    vegamsp.com

  • VegaNext.com

    Company site

    Site for VegaNext, an AI-native managed service provider for enterprise cybersecurity, AI automation and infrastructure management.

    veganext.com

  • MyLaffApp

    Web app

    Laughter wellness app that runs short guided laughter sessions.

    app.mylaffapp.com

From commit to monitored deploy

DevOps

CI/CD and observability pipeline

Ipact Consult, 2024

End-to-end delivery pipeline with quality gates and monitoring.

−40% build time−25% production bugs
  1. CommitGit
  2. JenkinsBuild and test
  3. SonarQubeQuality gate
  4. NexusArtifact storage
  5. DockerDeploy
  6. Grafana, PrometheusMonitoring

Stack: Jenkins, SonarQube, Nexus, Docker, Grafana, Prometheus, Spring Boot

Freelance and other projects

  • AdGen.tn Freelance

    AI tool for Tunisian sellers. From product photos, it checks the listing, reworks the visuals, films an intro shot and writes a sales page in French.

    AI
    adgen.tn

  • Karhabtii.tn Freelance

    Online store for car parts, tyres and accessories in Tunisia: 50,000+ parts with compatibility checked per vehicle, and delivery across the country.

    karhabtii.tn

  • IVPiecesAuto Freelance

    Online store for car parts, tyres and accessories in Tunisia, with 180,000+ parts and compatibility checked per vehicle.

    ivpiecesauto.com

  • Aquafun Marina Freelance, 2024

    Live booking platform for a water-sports marina (catalog, booking, customer management).

    React, Node.js, Express, MongoDB

    aquafunmarina.com

  • Zorraga Auto 2024

    E-commerce for Tunisia's #1 auto-parts distributor (catalog, stock, online orders).

    Angular, Spring Boot, PostgreSQL

    zorraga.com

  • CourZello 2024

    E-learning platform (Coursera-style) with AI course recommendations, certifications and content management.

    Angular, Spring Boot, Flask, MongoDB, REST API

Academic projects also include Trippee and Smart Center.

Skills

AI and LLMs

  • LLM API integration
  • RAG
  • Chatbots
  • AI agents
  • Voice AI (speech-to-text)
  • NLP
  • Machine Learning
  • Deep Learning
  • Transformers
  • N8N automation

DevOps and cloud

  • Docker
  • Docker Compose
  • Linux servers (VPS)
  • Nginx
  • SSL/TLS
  • CI/CD pipelines
  • Jenkins
  • SonarQube
  • Nexus
  • Grafana
  • Prometheus
  • AWS

Back-end

  • Python
  • FastAPI
  • Flask
  • Node.js
  • Express.js
  • Spring Boot
  • Django
  • ASP.NET
  • REST APIs
  • Microservices

Front-end

  • React.js
  • Next.js
  • TypeScript
  • JavaScript
  • Angular
  • HTML5
  • CSS3

Data

  • PostgreSQL
  • MongoDB
  • MySQL
  • PL/SQL

Programming languages

  • Python
  • TypeScript
  • JavaScript
  • Java
  • SQL
  • C#
  • C++
  • R

Tools and practices

  • Git/GitHub
  • Jira
  • Postman
  • VS Code
  • Agile/Scrum
  • TDD

Certifications

Let's talk.

I'm open to AI Engineer, MLOps and DevOps roles, remote or with relocation, nationally or abroad.

Emailaymen.rahali@esprit.tn Phone+216 92 554 097 LinkedInlinkedin.com/in/aymen-rahali GitHubgithub.com/Rahali-Aymen CV, EnglishCV_Aymen_Rahali_AI_DevOps_EN.pdf CV, FrenchCV_Aymen_Rahali_AI_DevOps_FR.pdf CV, ArabicCV_Aymen_Rahali_AI_DevOps_AR.pdf

Download my CV

Which language would you like it in?

EnglishEnglish version, PDF FrançaisFrench version, PDF العربيةArabic version, PDF