Freelancer profile translated to English.

Description

I am a Lead Data Scientist & Fullstack Engineer, specializing in transforming data into concrete business levers. My approach goes beyond code: I design and deploy solutions that maximize ROI, margins, and operational efficiency.

---

🚀 What I do

Retail Media & Marketing: incrementality measurement, attribution, ROAS optimization, negotiation support between brands and retailers.

Data Engineering & Cloud: scalable pipelines (GCP, AWS), BigQuery ingestion, reporting automation (Pub/Sub, Cloud Functions, Cloud Run).

Application Development: interactive dashboards, custom CRM (FastAPI + React), marketplace connectors (PrestaShop ↔ Zalando, Amazon, Spartoo).

Artificial Intelligence: LLM integration (GPT-4o, Ollama), conversational agents, automatic insight generation, explainable predictive models (PyTorch + SHAP).

---

🎯 My Approach

1. Business-first vision: every deliverable is designed to generate a competitive advantage, not just to run in production.

2. Industrialization: I transform POCs into robust, secure, and scalable products.

3. User simplicity: intuitive interfaces, reliable automation, measurable time savings.

4. Trust & transparency: clear communication, on-time delivery, comprehensive documentation.

🔧 Technical Stack

Cloud: GCP (BigQuery, Cloud Run, Functions, Pub/Sub, GCS), AWS (Lambda, Elasticsearch), Azure DevOps

Backend: Python (FastAPI, pipelines, APIs), Node.js

Frontend: React / Vite.js, interactive dashboards

Data Science: PyTorch, NLP, Generative AI, SHAP (explainability), Q-learning / MCTS

DevOps: Docker, CI/CD, Makefile automation, IAM/IAP, web app security

Business Tools: Google Sheets + App Script, PrestaShop, marketplace integrations

Languages

French
Native or bilingual
English
Fluent

Workplace preferences

Remote only

Primarily works remotely

Galec
Lead Data Scientist
DIGITAL AND IT
April 2024 - Today (2 years and 2 months)
Paris, France
Infrastructures & Pipelines
GCP Migration and Industrialization: design of BigQuery + Cloud Functions + Cloud Run pipelines for retail media data ingestion and processing (volume > 40K lines / file).

Automation: weekly/monthly scheduling via Cloud Scheduler + Pub/Sub, secure storage on Cloud Storage, job monitoring.

Performance Optimization: parallelization via ThreadPoolExecutor, reduction of processing times from several minutes to a few seconds.

Modeling & AI
Attribution & Incrementality: implementation of econometric models to measure the incremental ROAS of media campaigns.

AI-Augmented Reporting: integration of GPT/LLMs to generate automatic comments on reports (results, trends, recommendations).

Intelligent Indexing: creation of contextual indexes to query documents (PDF, BigQuery, Sheets) in natural language.

🖥Products & Applications
Retail Media Dashboard: development of React + FastAPI applications for real-time performance visualization (dynamic filters, customer segments, campaigns, margins).

Conversational AI Agent: integration of GPT-4o into a dashboard to translate business queries into backend API calls (Compute KPIs, BigQuery queries).

📈 Reporting & Activation
BRD Reports: automation of monthly & quarterly report generation, with storage in BigQuery + GCS and CSV/XLSX export.

Advanced Visualizations: heatmaps, radar charts, real-time simulations with Faker data.

👥 Leadership & Methodology
Data Team Structuring: implementation of standards (Makefile, unit tests, CI/CD, GCP monitoring).

Skill Development: internal training on securing exposed web applications and best practices for cloud/data engineering.

Strategic Vision: ROI-oriented framing, margins, efficiency to align data science and commercial performance.
Google Cloud Cloud GCP Artificial intelligence MLOps Web App
Fortil
DATA ENGINEER
DIGITAL AND IT
May 2021 - March 2023 (1 year and 10 months)
Paris, France
Softbank Robotics
Design of a serverless API (AWS Lambda + Elasticsearch) to centralize health metrics for thousands of robots deployed in real-world conditions.

Implementation of a real-time log pipeline with data ingestion, indexing, and visualization (Kibana).

Development of proactive anomaly detection mechanisms (health scores, dynamic thresholds).

Implementation of intelligent alerting for preventive maintenance.

AccorHotel
🏨 AccorHotels – Dathomir

Mission: build a centralized data platform to leverage customer data potential and improve personalization of the hotel experience.
Software Architecture Data Engineer Python Reporting DevOps
Zalando
Software Engineer
RETAIL (SMALL BUSINESS)
September 2020 - Today (5 years and 9 months)
Berlin, Germany
🛍 Zalando – Marketplace Integration

Mission: connect PrestaShop to Zalando APIs to automate catalog and order management, in a multi-channel e-commerce expansion strategy.

🔧 Technical Achievements

Development of a PrestaShop ↔ Zalando API connector:

automated product feed generation and updates (stock, prices, attributes).

real-time order & return synchronization via webhooks.

Implementation of an update scheduler (CRON/Cloud jobs) to ensure feed reliability.

Creation of an admin dashboard (FastAPI + React) to:

monitor feeds (upload success/errors).

manage stock and orders.

handle exceptions with real-time alerts.

Securing the integration via OAuth2, JWT, centralized logs.

🎯 Business Impact

Reduction of catalog and order management time from several hours/day to almost zero manual intervention.

Increased feed reliability: decrease in upload/return errors thanks to upstream automatic validation.

Improved multi-marketplace expansion capability (Zalando, Spartoo, Cdiscount, Amazon) from a unified technical foundation.

Direct ROI: accelerated time-to-market for products and maximized sales on Zalando.
Project management Web marketing strategy API Data visualization Customer service