Hard Skills — Data, AI & Architecture

I work on data structuring and architecture choices that keep products robust, performant and evolvable over time. On AI topics, my approach is pragmatic: AI should solve a real business problem, not just follow a trend.

Core capabilities

Databases

PostgreSQL, MongoDB, MariaDB, Redis, relational and NoSQL modeling, query optimization

Applied AI

RAG (Retrieval-Augmented Generation), LLM integration, data pipelines, intelligent automation

Architecture

Solution Architecture, service decomposition, API design, hexagonal architecture, scalability

AI and RAG: my positioning

RAG is currently one of the most effective approaches to integrate AI into real business applications. Instead of costly model fine-tuning, the context of an LLM is dynamically enriched with relevant data from the system (knowledge base, documents, historical records).

Design of RAG pipelines (ingestion, vectorization, retrieval, generation).
Integration of LLMs (OpenAI, Anthropic, open-source models) into existing backend architectures.
Automation of repetitive workflows with AI serving concrete business operations.
Evaluation of generated answer relevance and reliability before production rollout.

Software architecture: applied principles

Data schema design that remains readable, consistent and evolvable with the business domain.
Context-driven architecture choices based on team size, product maturity and load constraints.
Explicit technical trade-offs: maintenance cost, migration risk, accidental vs essential complexity.
API design: REST conventions, naming standards, versioning and automated docs (Swagger/OpenAPI).
Patterns: Repository, lightweight CQRS, layered boundaries, pragmatic SOLID usage.

Related experience

Apogée Consult · Co-Founder & CTO (Jul 2025 – Present)

Design and integration of AI building blocks (RAG, automation) in custom digital products for B2B clients. Definition of application architectures: database choices, service decomposition and integration strategy.

Métropole de Lyon & Superwyze (2023 – 2025)

Strong hands-on work on data modeling (PostgreSQL, MariaDB, MongoDB), third-party API integration and data-access layer structuring in real production contexts.