Portfolio / AI · ML Engineering

Pol QUIMERC'H.

Work-study Master's student, I build Machine Learning systems with concrete, measurable impact.

Portrait of Pol QUIMERC'H in profile, looking upward
locBrest → Lyon, France
seekingPermanent · R&D / Data Sci.
fromOctober 2026
01 · about

I started out in full-stack development before moving into AI engineering. I hold a professional bachelor's degree and am now specialising in Machine Learning and Deep Learning through a work-study Master's (RNCP level 7) at EPSI. Alongside it, I am completing three years as a data / ML apprentice at Eureden, an agricultural cooperative.

I work on applied AI and care about the engineering behind it, from exploring the data to running a reliable system in production. I am looking for a permanent position in Lyon, France, in R&D, Data Science or ML engineering, available from October 2026.

02 · selected projects
From data to production
001
National electricity demand forecasting

An open-source MLOps project that takes the model from a notebook to a FastAPI service on AWS, with automated retraining, drift detection (PSI) and Prometheus monitoring.

MAPE 2.37% · R² 0.967
002
MilkCast, milk collection forecasting

At Eureden, I industrialised a critical manual process with a serverless data pipeline (AWS Lambda, S3). It projects volumes over a rolling 36 months and supports scenario simulation.

36-month projection
003
Digicrop, crop yield from satellite imagery

I built a Sentinel-2 pipeline from scratch (STAC, COG partial reads, NDVI/EVI/NDWI) to enrich XGBoost yield models, then ran an ablation study to measure the gain.

R² +0.18

Full case studies, with charts and production screenshots, are on the French projects page.

03 · skills
Day-to-day toolkit
Machine LearningDeep LearningComputer Vision Data ScienceMLOpsPyTorchscikit-learn XGBoostTransformersOpenCVPandas NumPySHAPFastAPIDocker AWSGitPython

Prefer to ask directly? The assistant at the bottom right answers in English, about my background, projects or the role I am after.