Looking for a job

Mathieu BARBIER

@mathieubarbier

Message

ML engineer and BELYA co-founder building creator-content analysis and training PyTorch models at Incenteev.

Guadeloupe

Message

At Incenteev, I generated and validated training data, then rebuilt the labelling process with blind review and a long-tail holdout. I fine-tuned a PyTorch embedding encoder that beat a large commercial embedding API on unseen data.

At BELYA, where I'm co-founder and CTO, I develop an on-device content-analysis pipeline for live creator data, with stages benchmarked for quality, latency, and memory. I also build ML-debugging tasks for coding agents and have piloted them with a coding agent.

Experience

Work history, roles, and key accomplishments

Current

Co-founder & CTO

Current

BELYA

Mar 2026 - Present (7 months)

Co-founded and led technology for an influencer-marketing platform, building on-device content-analysis pipelines including ASR, voice separation, VLM keyframe descriptions, OCR, and niche classification systems.

ASR VLM OCR Embeddings Classification PyTorch Python Docker GitHub Actions

AI Project Lead

Incenteev

Feb 2026 - Jul 2026 (5 months)

Sole ML engineer for an end-of-studies internship, building training data pipelines, contrastive fine-tuning of Transformer embedding encoders, knowledge-tracing models, and evaluation harnesses for an adaptive sales-training platform.

PyTorch LLM HMM Baum Welch LSTM Evaluation Calibration Python

Education

Degrees, certifications, and relevant coursework

Grenoble INP

Master's degree (Diplôme d'Ingénieur), Computer Science & Applied Mathematics

2023 - 2026

Master's degree (Diplôme d'Ingénieur) in Computer Science and Applied Mathematics with a focus on Data Science and AI, including statistics, optimisation, and machine learning.