Skip to content

Available for AI projects

Hi, I'm

Mohamed SmaouiAI Engineer & Senior Full-Stack Developer

I build AI-First systems: autonomous agents, RAG pipelines and MCP servers, shipped to production with real software engineering discipline.

See my AI projectsGet in touch
Sfax, Tunisia GitHub LinkedIn
User / TriggerSupervisorRAG agentMCP agentVoice agentVector DBDB · APIsSTT · TTS

About

From full-stack to AI systems in production

Software engineer with 8 years of experience in software engineering and full-stack development, who has successfully moved into AI systems engineering (AI-First Systems) and intelligent automation.

I specialise in multi-agent architectures (LangGraph, LangChain), advanced RAG pipelines, MCP server development and LLM orchestration. I pair this AI expertise with strong software architecture skills (Python, Next.js, Vue 3, Laravel 12) and DevOps practices (Docker, Dokploy) to ship robust, scalable, production-ready generative AI solutions.

years of experience
8+
companies & missions
5
web projects shipped
10+
spoken languages
3

AI expertise

What I build with LLMs

AI building blocks designed for production: orchestrated, tool-enabled, observable and deployed.

  • Multi-agent architectures

    Designing and orchestrating multi-agent systems with LangGraph and LangChain: state graphs, supervisors, memory and tools to automate business workflows end to end.

    LangGraphLangChainPythonTool calling

  • RAG pipelines & contextual retrieval

    Ingestion, chunking, embeddings and optimised contextual retrieval to give LLMs reliable, up-to-date context.

    EmbeddingsVector DBReranking

  • MCP servers & tool integration

    Model Context Protocol servers that connect models to databases, third-party APIs and in-house tools.

    MCPTypeScriptPythonREST / GraphQL

  • Multimodal voice AI

    Speech-to-Text and Text-to-Speech integration for voice interfaces and conversational agents.

    STTTTSStreaming

  • AI deployment & DevOps

    Containerising and deploying scalable AI services and orchestration infrastructure.

    DockerDokployGitLab CI/CDLinux

  • Full-stack AI products

    Fast Next.js / React interfaces backed by Python microservices: AI shipped as a real product, accessible (WCAG / RGAA) and maintainable.

    Next.jsReactPythonLaravel 12Vue 3

AI projects

Artificial intelligence use cases

Representative examples of the systems I design. Client details are confidential.

01

Multi-agent business workflow orchestrator

Problem
Repetitive business processes spread across many tools and teams.
Solution
A LangGraph supervisor delegating to specialised agents (analysis, research, action), with human-in-the-loop validation at critical steps.
Architecture
  1. Trigger
  2. Supervisor
  3. Agents
  4. Tools
  5. Review
LangGraphLangChainPythonRedisDocker

02

RAG knowledge assistant

Problem
A large document base that is hard to query and quickly goes stale.
Solution
RAG pipeline with contextual retrieval, embeddings and cited answers, served in a streaming Next.js interface.
Architecture
  1. Docs
  2. Chunking
  3. Embeddings
  4. Retrieval
  5. LLM
PythonVector DBEmbeddingsNext.js

03

MCP server for internal data

Problem
LLMs have no access to the company's databases and in-house APIs.
Solution
A Model Context Protocol server exposing typed, secured tools (DB reads, API calls) usable from any MCP client.
Architecture
  1. LLM client
  2. MCP
  3. Tools
  4. DB / APIs
MCPTypeScriptNode.jsMySQLREST

04

Voice conversational agent

Problem
Users who need to interact by voice, in real time.
Solution
Streaming STT → LLM agent → TTS chain, handling interruptions and calling tools mid-conversation.
Architecture
  1. Voice
  2. STT
  3. Agent
  4. TTS
  5. Voice
STTTTSLangChainPythonWebSockets

Career

Professional experience

  1. AI Engineer & Senior Full-Stack Developer

    Zetabox · Tunisia

    April 2026 – Present

    Current role · AI role
    • Designed and orchestrated complex multi-agent systems with LangGraph and LangChain to autonomously automate business workflows.
    • Built advanced RAG pipelines and contextual retrieval architectures optimised for LLMs.
    • Developed MCP servers connecting models to databases, third-party APIs and in-house tools.
    • Integrated STT / TTS models for voice interfaces and intelligent conversational agents.
    • Containerised and deployed scalable AI services with Docker and Dokploy.
    • Next.js / React interfaces backed by Python microservices and processing scripts.
    PythonLangGraphLangChainMCPRAGNext.jsReactSTT/TTSDockerDokploy

  2. Senior Full-Stack Developer (Freelance)

    Orvea · France

    January 2026 – March 2026

    • Scalable web applications with Laravel 12 for the core and Vue 3 (Composition API) for the UI.
    • Applied SOLID principles and built modular, high-performance components.
    Laravel 12Vue 3MySQLREST APIsGit

  3. Front-End Developer

    Futur Africa · Sfax, Tunisia

    July 2025 – September 2025

    • Designed a modular, optimised and reusable Next.js / React boilerplate.
    • Integrated and consumed GraphQL services for flexible data handling.
    Next.jsReactGraphQLStorybookSCSSGitLab

  4. Support, Maintenance & Accessibility Engineer

    Stratis World Wide · Sfax, Tunisia

    December 2023 – June 2025

    • Technical audits and RGAA / WCAG compliance of application modules.
    • Evolutive maintenance of legacy and modern applications, mentoring junior developers.
    JavaScriptTypeScriptNunjucksSCSSGitLab

  5. Full-Stack Developer / Web Integrator

    Netnovus · Sfax, Tunisia

    October 2018 – July 2023

    • End-to-end design of a bike rental platform with Vue.js and Laravel.
    Vue.jsLaravelTailwind CSSBootstrapMaterial UIMySQL

Stack

Technical skills

AI & LLM orchestration

  • LangGraph
  • LangChain
  • Multi-agent systems
  • RAG
  • Contextual Retrieval
  • MCP
  • AI Tool Integration
  • Speech-to-Text
  • Text-to-Speech
  • Prompt engineering

Languages & back-end

  • Python
  • TypeScript
  • JavaScript
  • Node.js
  • Laravel 12
  • REST APIs
  • GraphQL

Front-end

  • Next.js
  • React
  • Vue 3
  • Alpine.js
  • Tailwind CSS
  • Material UI
  • SCSS

DevOps & AI infrastructure

  • Docker
  • Dokploy
  • GitLab CI/CD
  • GitHub
  • Git
  • Linux / WSL
  • Storybook

Databases & vector DBs

  • MySQL
  • MongoDB
  • Redis
  • Vector databases
  • Embeddings

Quality & methods

  • WCAG
  • RGAA
  • Technical audits
  • Software architecture
  • Scrum / Kanban
  • Mentoring

Web work

Shipped web projects

Before AI: web platforms designed and delivered in agency and freelance settings.

Education

Education & certifications

Certification

Anthropic CCAF training — Claude Certified Architect / Practitioner

Anthropic

Generative AI architectures, Claude model orchestration, advanced prompt engineering and agentic architectures.

Degree

Computer Engineering Degree — Web Development major

Esprit, Tunisie · 2018

Engineering degree in software engineering and web development.

Languages

  • Arabic — native
  • French — professional (B2)
  • English — professional (B2)

Interests

Sports, video games, gardening and DIY

Contact

Let's build your next AI system

Working on agents, RAG or an LLM integration? Let's talk.

Send an email