AI that ships,not demos that stall.
I build LLM agents, computer vision models and data pipelines that hold up in production: tested, measured and documented so your team can own them after I leave.
Lille, France · Remote worldwide

Services
Four ways I can move your project forward
Each engagement ends with working software, tests and documentation, never a slide deck alone.
LLM agents and MCP integration
Agents that call your tools and data safely, connect through MCP, and are evaluated before they reach users.
Deliverables
- MCP servers and tool integrations
- RAG over your documents
- Multi-agent workflows
- Evaluation and regression suites
Machine learning and computer vision
Models measured on the metric that matters for your problem, with a protocol that makes results reproducible.
Deliverables
- Segmentation and detection models
- Annotation tooling and datasets
- Benchmarks and model selection
- Inference packaging
Data pipelines and MLOps
Ingestion, APIs and monitoring that keep running after the demo, containerized and wired into CI.
Deliverables
- FastAPI services
- Time-series ingestion and forecasting
- Docker and CI/CD pipelines
- Monitoring with Prometheus
Selected work
Built, measured, delivered
Research tooling, agent platforms and data systems. Every figure below comes from the actual project.
Process
How a mission runs
- 1
Frame
A short call, then a written brief: the real problem, the success metric and what is out of scope. Numbers come before code.
- 2
Prototype
A working slice on your data within the first iteration, so we judge direction on evidence rather than opinions.
- 3
Harden
Tests, evaluation harness, CI and packaging. The goal is a system that behaves the same on Monday as it did in the demo.
- 4
Hand over
Documentation, a walkthrough session and clean ownership. You keep the code, the data and the know-how.
Ways to work together
Pick the format that fits the job
Every engagement starts with a free 20-minute call and a written estimate. No surprise invoices.
Fixed-scope project
For a clear deliverable: an agent, a model, a pipeline or a working prototype.
- Written brief and fixed estimate
- Agreed timeline and milestones
- Tests, documentation and handover included
Sprint or day rate
For teams that need extra engineering hands on an existing codebase.
- Works in your repo and tools
- Weekly progress notes
- Flexible length, billed per day
Audit and advice
For a second opinion before you commit to an AI or data architecture.
- Review of code, data and evaluation
- Written report with prioritized fixes
- Follow-up call to walk through it
About
Engineer first, with a research habit
I am an engineering student in Data Science and Responsible AI at IMT Nord Europe, and I am completing a second engineering degree in computer science at AIAC in Casablanca.
At Oracle I worked inside the open-source runtime behind Agent Spec: fixing runtime defects, adding MCP support and helping product teams ship their agents. At CERI MP I automated a scientific imaging task end to end, from the annotation tool to a packaged desktop application.
What carries across: I measure before I optimize, I write the tests that catch silent failures, and I leave behind documentation people actually read.
Certifications
- OCI 2024 Certified AI Foundations AssociateOracle
- OCI 2025 Certified Foundations AssociateOracle
- Deep Learning SpecializationCoursera
- Machine Learning SpecializationCoursera
- Building Generative AI Skills for Developers
- DevOps Foundations
Experience
May to Sept 2026 · Dunkerque, France
CERI MP, IMT Nord Europe
Research Engineer Intern, Computer Vision
Automated dislocation extraction from in-situ electron microscopy films: annotation tool, 890-pair dataset, controlled model benchmark, CPU-only instance separation library and a packaged desktop application.
July to Sept 2025 · Casablanca, Morocco
Oracle
AI Integration Engineer Intern
Contributed to Wayflow, the open-source runtime for Agent Spec: fixed 5 critical runtime defects, added MCP support, refactored Oracle Database integration and maintained the Sphinx documentation.
July to Aug 2024 · Casablanca, Morocco
Arrow Electronics
DevOps and Integration Intern
Stood up the deployment environment of a Spring Boot microservices order system with Kafka, MySQL and Docker Compose, and wired Postman integration tests into Jenkins.
Education
2025 to 2027 · Lille, France
IMT Nord Europe
Engineer's Degree in Data Science and Responsible Artificial Intelligence
2023 to 2027 · Casablanca, Morocco
Académie Internationale de l'Aviation Civile
State Engineer's Degree in Computer Science
Contact
Let's talk about your project
Share a few lines about what you are building. I answer every message personally, usually within 48 hours.
Or reach me directly