Description

I build computer vision and AI systems that work in the real world — not just in notebooks.

AI & Computer Vision Engineer based in Luxembourg with an MSc in Computer Vision & AI (University of Luxembourg) and 2.5 years of industrial experience deploying real-time visual inspection systems on live production lines. When a defect-detection model has to survive vibration, changing light and 24/7 line speed, "it works on my laptop" isn't good enough — that's the standard I bring to every project.

What I deliver:

• Custom object detection & segmentation models (YOLO11, SAM 2) trained on your data

• Industrial visual inspection: camera/optics/lighting selection + algorithm + deployment

• AI chatbots (Claude/GPT) grounded in your business knowledge base

• LLM workflow automation (n8n, Make, Zapier) with deterministic checks around the LLM

• Dataset annotation pipelines and ML evaluation

Proof of work (open source, live demos):

• Reproducible detection/segmentation benchmark (YOLO11, RF-DETR, SAM 2, Mask2Former) — Zenodo DOI

• Building-intelligence RAG copilot over public data from 3 countries

→ github.com/andrudebaran7

Languages: Spanish (native), English (fluent)

Industry field of expertise

Languages

English
Fluent
Spanish
Native or bilingual
French
Basic
German
Basic

Workplace preferences

Remote only

Primarily works remotely

Freelance
Independent AI Engineer — Applied AI & Automation
September 2025 - Today (10 months)
Built an open benchmark of object-detection and segmentation models for industrial use cases (PyTorch/YOLO) with a live demo and Zenodo DOI — github.com/andrudebaran7.
Built a building-intelligence RAG copilot over public data from 3 countries, with evaluated PDF extraction and a live Streamlit demo.
Designing LLM-based customer-service assistants (Claude/GPT) and workflow automation for SMEs (reservations, customer communication).
LLM YOLO Pytorch
Mesolt Inspect
Machine Vision Engineer
April 2023 - August 2025 (2 years and 4 months)
Luxembourg
Designed and deployed real-time visual inspection systems on live production lines, replacing manual quality control.
Analysed client requirements (inspection speed, rejection rate) and selected the full hardware chain: 3D, matrix and line-scan cameras, optics and lighting.
Developed and validated detection algorithms (Python, OpenCV, deep learning; Keyence platforms), ensuring compliance with industrial quality standards.
Collaborated with engineering, production and quality teams and drove continuous performance optimisation.
Computer Vision Python Deep Learning OpenCV
University of Luxembourg
Research Assistant (Student Employee)
November 2021 - March 2022 (4 months)
Luxembourg
Supported research on manufacturing-process automation using computer vision and AI; developed and tested vision-based solutions for industrial applications (Python, MATLAB, TensorFlow, PyTorch, OpenCV).