Professional Summary
Freelance Senior Data Scientist with 12+ years of industry experience and 4+ years of PhD research. Currently building agentic pipelines and evaluation frameworks for enterprise data (AstraZeneca), and open to new data science engagements. My expertise spans building discrete event simulations for network optimization (Glovo), implementing vehicle routing problems (VRP), and designing evaluation frameworks for LLM agents (Meight).
Last updated: August 2026 · ⬇ Download PDF (recruiter version)
Core Skills
Languages:
Python, C++, SQL · Go (working knowledge)
ML/AI:
LLM Evaluation, Agent Systems, RAG, Knowledge Graphs, Machine Learning, Deep Learning
Optimization:
Discrete Event Simulation, Vehicle Routing (VRP), Discrete Optimization
Tools:
PyTorch, LangGraph, LangSmith, Google OR-Tools, DeepEval
Dev Tools:
GitHub Copilot, Cursor, VS Code, bash/zsh, AWS
Connect
💡 Tip: The PDF version is clean and ready to share. The recruiter-optimized version is ATS-friendly, 1-2 pages, and focused on recent experience.
📍 Location: Tiana (Barcelona area), Catalonia, Spain
💼 Work Setup: Freelance · Remote
🗣️ Languages: Dutch (native), English (fluent), Spanish (fluent), Catalan (proficient)
🇪🇺 Citizenship: Dutch / EU citizen
* Photo edited with AI (Gemini) for background removal and minor cleanup
Experience
Artificial Intelligence Engineer
AstraZeneca (contract via Hybrid Mind) · Freelance · Mar 2026 - Present
Barcelona, Spain · Remote
- Agentic Pipelines: Design and build LLM agent pipelines for enterprise data workflows with LangGraph.
- Agent Evaluation: Define evaluation suites for agent reliability — accuracy, faithfulness, tool use, cost — using LangSmith tracing and offline/online evals.
- Semantic Modelling: Model a semantic layer and domain ontology / knowledge graph over enterprise data so agents query modelled entities and relationships rather than raw tables.
Skills: Python, LangGraph, LangSmith, LLM Evaluation, Agent Systems, Knowledge Graphs, Semantic Layer
Senior Data Scientist
Meight · Dec 2024 - Feb 2026 · 1 yr 2 mo
Évora, Portugal · Remote
- LLM Agents: Designed the company's agent evaluation framework, combining classical ML metrics with LLM-as-a-judge evaluation (faithfulness, cost) to validate agent reliability before production.
- Tour Optimization: Developed freight planning algorithms that integrate real-time telematics and EU regulatory compliance (Reg. 561/2006) to optimize driver schedules.
Skills: Python, Go, LLM Evaluation (DeepEval), Agent Systems, Route Optimization
Senior Data Scientist
Glovo · Nov 2019 - Nov 2024 · 5 yrs
Barcelona, Spain · Hybrid
- Simulation Leadership: Led the architecture and refactor of the discrete event simulator for Glovo's delivery network (23 countries, 120,000+ couriers). The simulator modelled the full delivery lifecycle, enabling the team to test new features and run hyperparameter optimization before production.
- Route Optimization: Led the VRP implementation for "Live Multibundling," enabling dynamic order batching in real-time.
- Matching Logic: Built the automated optimization pipeline for the matching cost function, enabling rapid parameter tuning for network efficiency.
Skills: Python, Discrete Event Simulation, Discrete Optimization, Vehicle Routing (VRP)
Senior R&D Software Developer
Atomian · Jun 2017 - Nov 2019
Sant Cugat, Spain
- Data Engineering: Built ETL systems to prepare and integrate data for Atomian's cognitive architecture, enabling the Medical Records product to extract structured information from unstructured patient reports.
- Document Processing: Developed pipelines to read and process legal/property documents (notas simples) and medical records using Atomian's NLP platform.
- Analytics & Prediction: Performed analytics and built prediction models on extracted healthcare data, ensuring accuracy and reliability in clinical contexts.
Skills: C++, Information Extraction, Cognitive Architectures, Data Analytics
PhD Researcher
IRI - Institut de Robòtica i Informàtica Industrial (CSIC-UPC) · Apr 2011 - Jun 2017 · 6 yrs 3 mos
Barcelona, Spain
- Thesis: Searching and Tracking of Humans in Urban Environments by Humanoid Robots.
- Developed algorithms for robots to detect and track people using POMCP, particle filters, and Reinforcement Learning.
- Validated methods through 3km+ of autonomous experiments with humanoid robots in real-world urban settings.
Skills: C++, Python, ROS, POMCP, Particle Filters, Robotics
Software Engineer
Vestas Eólica SAU (Wind Energy) · Jan 2008 - Feb 2011 · 3 yrs 2 mos
Barcelona & Madrid, Spain
- ERP Development: Designed and developed an ERP system to manage, plan, and optimize maintenance operations for wind turbine fleets.
- Performance & Diagnostics: Worked in the P&D department performing statistical error analyses of wind turbines and developing KPI/performance dashboards. Led the technical implementation of a 3-month test to statistically validate wind and turbine production forecasts from weather companies (database design, statistical comparisons including t-tests).
- SAP Migration: Supported the migration of legacy ERP systems to SAP enterprise platform.
Skills: Java, SQL Server, MS Access, Excel VBA, ERP Systems, Statistical Analysis
Early Career
Software engineering and research roles during and shortly after university (2001 - 2008).
- Master's Thesis Researcher - IIIA-CSIC, Barcelona · 2007 · Bio-inspired robot navigation (Average Landmark Vector) and computer-vision landmark extraction from panoramic images. Thesis (PDF)
- Software Engineering Intern - Vestas Eólica SAU, Barcelona · 2007 · Document digitalization, database automation, and KPI dashboards for the service department.
- Software Engineer - DevSquad, Netherlands · 2004 - 2006 (part-time) · Integration components between laboratory behavioral-analysis and sensor-data systems; GLP digital-signature module.
- Software Engineer, Industrial Placement - Thales Nederland BV, Hengelo · 2003 · Ported the Man-Machine Interface of the IRScan infrared targeting system (Goalkeeper anti-missile defense) from Sun Sparc to PC Linux. Report (PDF)
- Junior Software Engineer, Orientation Internship - University of Groningen, Legal Informatics · 2001 - 2002 · Java database tools and an RMI client/server application for a legal expert system; non-monotonic logic operators. Report (PDF, Dutch)
Skills: Java, C#, C, C++, Matlab, RMI, Computer Vision, Expert Systems, Robot Navigation
Education
Doctor of Philosophy (PhD) - Automatic Control, Robotics and Computer Vision
Universitat Politècnica de Catalunya · 2011 - 2017
Thesis: Searching and Tracking of Humans in Urban Environments by Humanoid Robots
Thesis PDF | Thesis Page
Master of Science (MS) - Artificial Intelligence
University of Groningen · 2005 - 2008
Thesis: Solving Ambiguity in Global Localization
Thesis PDF
Bachelor of Science (BS) - Computing Science
University of Groningen · 2003 - 2006
Thesis: SLAM using Growing Neural Gas
Thesis PDF
Bachelor of ICT - Information and Communication Technology
Hanzehogeschool Groningen (University of Applied Sciences) · 1999 - 2003
Graduation Project: IRScan MMI Platform Porting (PDF)
Academic Service
Peer Reviewer — IEEE/RSJ IROS 2026 (International Conference on Intelligent Robots and Systems)
Topic: Multi-Robot Coverage Planning · Pittsburgh, PA · Oct 2026
Certificates & Courses
DeepLearning.AI
PyTorch for Deep Learning · Feb 2026
3-course professional certificate
Machine Learning in Production · Mar 2026
Agentic AI · Nov 2025
Generative AI for Software Development · Feb 2026
3-course professional certificate
AI for Medicine · Jun 2026
3-course professional certificate
Show individual courses
Short Courses · Nov 2025 - Aug 2026 · 46 courses completed
LangChain Academy
LangChain & LangSmith Courses · Mar 2026
7 courses
Anthropic Education
Claude Courses · Apr 2026
8 courses
Coursera
Deep Learning Specialization · Jul 2019
5-course specialization by Andrew Ng (DeepLearning.AI)
Discrete Optimization · Dec 2020
Georgia Institute of Technology
View Certificate · Download PDF · Discrete Optimization Exercises
Coding Roots
Started programming young with QBasic and Turbo Pascal, creating my first programs. During my HBO study (HIO), I learned Java and C. My university Master in AI introduced me to Prolog, Haskell, Matlab, and R for AI/ML work. Throughout my PhD I worked extensively with C++, and from my PhD onwards I've primarily used Python for data science and optimization. Today, I increasingly use LLMs (GitHub Copilot, Cursor) to help build the main components, accelerating development while maintaining code quality.