Background

April 2023 → September 2026

Data/AI Engineer, apprenticeship at INSEI, Suresnes

• Production RAG and LLM architecture: a document assistant answering staff questions on internal procedures, contacts and resources, serving two sites across fourteen containerised services.
• Hybrid retrieval: embeddings, a vector store, keyword search and reranking. Moving from purely semantic to hybrid retrieval is what improved retrieval, and it was measured.
• Pipelines and MLOps: incremental ingestion, chunking, indexing, a REST API, containerisation and production rollout. Everything runs on the institution's own infrastructure, with no licence budget.
• Evaluation and observability: a versioned set of eighty-eight reference questions, an automated judge, and request tracing. That harness is what makes it possible to compare two pipeline versions instead of trading impressions.

See the project

May → July 2025

Full-stack developer, internship at Ace English Malta, in Malta

• Development in .NET and Angular within a three-person IT team, on a management, booking and reporting application: activities, payments, invoices, classrooms, courses and dashboards. Modules delivered end to end, from back-end to interface, plus the migration of a rich-text editor from one major version to the next across the existing application.
• International collaboration: team planning and tracking, version control, and secured REST APIs on the server side.
• The migration was carried out by leaving the old code path in place, disabled but intact, until the new one was fully validated, with a tracking file recording, screen by screen, what had been handled and what remained.
• In parallel, and without being asked, I catalogued twenty functional defects in parts of the application I had not written: missing input validation, actions still available after an item was deleted, success messages rendered in error colours, date inconsistencies. Most were fixed, each fix traceable in the repository history.
• The end-of-internship assessment, signed by the company, awards an A grade and singles out that initiative and the analytical work as strengths. It also states one area for improvement, which I reproduce as written: refining the accuracy of my time estimates on complex tasks.

2023 → 2026

Ingénieur en informatique, spécialité Data Science / IA, CESI École d’Ingénieurs, campus de Nanterre

Apprenticeship track.

2021 → 2023

Integrated two-year preparatory programme, CESI École d’Ingénieurs, campus de Nanterre