Berlin, Germany

BenjaminMoreno Torres

Data Science & AI Lead

About

Professional portrait of Benjamin Moreno Torres

I started as a mathematician. Six years studying the structure of things - probability, logic, the language underneath everything. Then, almost as a detour, I spent five years as an architect. Not because the math stopped mattering, but because I wanted to work with systems you could touch.

Data science found me somewhere between those two worlds. When machine learning started reshaping materials research, structural monitoring, and industrial R&D, I had something rare: the quantitative rigor to model it properly and the physical intuition to know when the model was lying.

For the past ten years I have been applying that combination - Bayesian Networks, Markov Chains, ontology engineering, NLP - to hard R&D problems: predicting material durability, building FAIR data pipelines, leading a Horizon Europe work package across six research institutions. The work has always been at the edge of what ML can do when the domain is genuinely complex.

I am now looking for a Data Science or AI Lead role in an organization with serious R&D - somewhere I can lead a team, own a data strategy, and work on problems that require both technical depth and the ability to communicate across disciplines. Berlin preferred, open to hybrid or remote.

Experience

April 2023 - May 2026

Work Package Leader and Probabilistic Life Cycle Models

Technische Universitat Berlin · Berlin, Germany

  • Led a work package spanning 6 partner organizations across 4 countries, owning deliverables, timelines, and technical integration end-to-end.
  • Applied Bayesian Networks and Markov Chains for predictive life-cycle assessment, enabling data-driven material selection decisions.
  • Designed ontology-based data architecture (OWL/RDF/SPARQL) for FAIR interoperability across 12+ heterogeneous research datasets.
  • Co-authored and coordinated industrial deployment of SLAMD - open-source AI tool for sustainable materials design, published in NanoWorld Journal (2023).
  • Supervised 4 graduate researchers in AI and digital construction.
  • Managed stakeholder communication across academia, industry, and EU project officers.
Bayesian NetworksHorizon EuropeOntology EngineeringPythonFAIR Data

2020 - January 2023

Machine Learning Researcher

BAM - Federal Institute for Materials Research · Berlin, Germany

  • Built predictive ML models (clustering, regression, ensembles) for material durability and circularity assessment.
  • Structured and analyzed databases on Alkali-Activated Concretes from heterogeneous experimental sources.
  • Developed data pipelines for EU digitalization projects, improving cross-institutional data quality.
  • Collaborated with chemists, physicists, and engineers to translate domain expertise into ML feature engineering.
Machine LearningMaterials SciencePythonData Pipelines

2016 - 2020

Data Scientist & ML Engineer

3S'TECH S.L. · Girona, Spain

  • Developed predictive algorithms for real-time structural health monitoring using IoT sensor networks.
  • Designed end-to-end data pipelines for sensor data ingestion, cleaning, transformation, and anomaly detection.
  • Implemented ML models for engineering databases and concrete testing.
IoTStructural Health MonitoringML EngineeringPython

2012 - 2016

Data Scientist - NLP & Analytics

Symanto GmbH · Nurnberg, Germany

  • Built Python and C# tools for social media sentiment analysis and market research.
  • Contributed to NLP pipelines (text classification, entity extraction) and graph-based influencer analytics.
  • Delivered data insights for digital marketing campaigns.
NLPPythonC#Graph Analytics

2005 - 2011

Structural Designer & Architectural Engineer

Various Firms · Spain

  • Structural design, planning, and valuation across residential, heritage, and public projects including the Disseny Hub Barcelona.
  • Developed systems-thinking and spatial reasoning skills that now underpin a distinctive approach to data architecture and ML system design.
ArchitectureStructural EngineeringSystems Thinking

Featured Project

SLAMD - Sequential Learning App for Materials Discovery

Open-source AI tool for inverse design of sustainable cementitious materials

SLAMD applies sequential (active) learning to accelerate the discovery of sustainable cement and concrete formulations. By iteratively selecting the most informative experiments, it reduces the number of lab tests needed to find optimal material compositions - making R&D faster and more sustainable.

Role: Co-developer, tester, and co-author. Contributed to design, validation, and the NanoWorld Journal publication (2023).

Sequential LearningActive LearningMaterials SciencePythonOpen Source

Skills

ML / AI

  • Bayesian Networks
  • Markov Chains
  • LSTMs
  • NLP
  • Graph Analytics
  • Clustering
  • Regression
  • Ensemble Methods
  • Sequential Learning

Programming

  • Python (primary)
  • R
  • C#
  • SQL

Data & Semantic Web

  • Ontologies (OWL, RDF, SPARQL)
  • FAIR Data Principles
  • Semantic Web
  • Linked Data
  • Data Pipeline Design

Leadership & Project Management

  • Horizon Europe
  • Multi-stakeholder coordination
  • International consortium management
  • Graduate researcher supervision
  • Technology transfer

Publications

First-author publication

An Ontology-Based Approach to Enable Data-Driven Research in the Field of NDT in Civil Engineering

MDPI Remote Sensing

View publication ->
First-author publication

Concreting a sustainable future: A dataset of alkali-activated concrete and its properties

ScienceDirect

View publication ->
First-author publication

Reincarnate D1.3 - Probabilistic methods for lifecycle assessment and prediction of buildings and construction products

Zenodo

View publication ->
Academic publication

Presenting SLAMD - A Sequential Learning Based Software for the Inverse Design of Sustainable Cementitious Materials

NanoWorld Journal, 2023

View publication ->
Academic publication

Data driven design of alkali-activated concrete using sequential learning

ScienceDirect

View publication ->
Academic publication

The Intersection Between Semantic Web and Materials Science

Advanced Intelligent Systems

View publication ->
Academic publication

A Perspective on Digital Knowledge Representation in Materials Science and Engineering

Advanced Engineering Materials

View publication ->
Academic publication

Meta-Learning for Adaptive Mix Design of Alkali-Activated Concrete

Springer

View publication ->
Academic publication

An Adaptive Upscaling Approach for Assessing Materials' Circularity Potential with Non-destructive Testing (NDT)

Springer

View publication ->
Academic publication

REINCARNATE: Shaping a Sustainable Future in Construction Through Digital Innovation

Springer

View publication ->
Academic publication

A case study of structural monitoring as a control tool in the restoration process of Heritage Structures: The strengthening of the Vistabella Church's Tower

UPCommons

View publication ->

Education

PhD, Semantic Web Applications for Materials Discovery

Technische Universitat Berlin

Ongoing

Master in Architecture (375 ECTS)

Universitat de Girona

2010

Master in Mathematics for Industry (60 ECTS)

Universitat Autonoma de Barcelona

2001

Degree in Mathematics (300 ECTS)

Universitat Autonoma de Barcelona

1994-2000

Contact

Open to conversations about Data Science and AI Lead roles in Berlin and the DACH region. Hybrid or remote.

benji.moreno@gmail.comConnect on LinkedIn