Zain Ul Abideen

Scientific machine learning researcher · BCMaterials, Leioa · PhD candidate, Computer Engineering, UPV/EHU

zainul.abideen@bcmaterials.net github.com/Enumk · Google Scholar (583 citations, h-index 15) · linkedin.com/in/zain-ul-abideen · ORCID 0000-0001-9909-2061

Interests

Multidisciplinary AI for scientific discovery.

  • Scientific ML for materials and chemistry: Scientific machine learning, machine learning interatomic potentials, quantum-dot simulation and nanomaterials modelling, computational chemistry and molecular modelling.
  • AI for life sciences: AI for drug discovery and cheminformatics, bioinformatics and computational biology.
  • NLP, generative and agentic AI: Natural language processing, generative AI and large language models, retrieval-augmented generation, agentic AI and autonomous systems.
  • ML systems and infrastructure: Distributed ML training and HPC model execution, reproducible ML workflows, model benchmarking and evaluation.

Education

  • PhD, Computer Engineering, University of the Basque Country (UPV/EHU), 2023–present, defence expected March 2027. Thesis: A Generic Automated and Adaptive Machine Learning Platform for Advanced Materials Research. Advisor: Prof. Ivan Infante. Funded by the IKUR Strategy (IKUR-HPC&AI).
  • Master of Science, Computer Science, COMSATS University Islamabad (CUI), 2017–2020, CGPA 3.46. Thesis: A Machine Learning Assisted Parts of Speech Tagger for Roman Urdu.
  • Bachelor of Science, Software Engineering, COMSATS University Islamabad (CUI), 2013–2017.

Research software

  • Orchestr.AI. Open-source platform that trains, fine-tunes, benchmarks and deploys 7 machine-learning interatomic potential architectures (SchNet, PaiNN, SO3net, FieldSchNet, NequIP, Allegro, MACE) from one YAML file, with SLURM and multi-GPU training on HPC. Release v0.2.0 on Zenodo.
  • QDSpace web platform. Quantum-dot builder and a library of DFT-relaxed structures with interactive PDOS and COOP visualisation, at quantumdotspace.org.
  • QDSpace Desktop. The QDSpace quantum-dot tools as a standalone desktop application, with a built-in local property calculator based on MACE.
  • Roman Urdu POS tagger (master’s thesis): LSTM-based tagger, new corpora, standardisation, tagset, and phonetic hashing.

Experience

  • Researcher, BCMaterials, Leioa, July 2023–present. Computational Materials Science Group, Prof. Ivan Infante; IKUR-HPC&AI project HPCAI11 ARK (2025–2026). Machine-learning interatomic potentials for quantum dots, Orchestr.AI and QDSpace, and co-development of a foundation-model interatomic potential for quantum dots.
  • Computer engineer, Ghulam Ishaq Khan Institute, Topi, November 2021–June 2023. AI researcher and lab instructor for six undergraduate AI courses.
  • Collaboration with research labs, November 2020–May 2023. Jeju anomaly detection, and MiLab metabolomics at Universitat Rovira i Virgili.
  • NLP engineer, Fiverr, Level 1, February 2020–July 2023. Client work in natural language processing.
  • Team lead, NLP Lab (Volunteer), COMSATS, January 2020–November 2021. Voluntary student role; a team of about 20.
  • Research associate, COMSATS, February 2018–November 2021. Text mining, natural language processing, and machine learning.

Selected publications

  • In preparation. Orchestr.AI: A Unified Workflow for Training and Benchmarking of Machine-Learned Interatomic Potentials.
  • 2026. Orchestr.AI v0.2.0: Unified Platform for Machine-Learning Force Fields. Zenodo software release.
  • 2026. Benchmarking Machine-Learning Interatomic Potentials for Dynamical Stability in Inorganic Semiconductor Nanocrystals: A CdSe Case Study. arXiv:2609.15299 (preprint).
  • 2024. A Novel NLP-Based Stock Market Price Prediction and Risk Analysis Framework. Journal of Computer Science and Technology.
  • 2023. Towards Effective Emotion Detection: A Comprehensive Machine Learning Approach on EEG Signals. BioMedInformatics.
  • 2022. Significance of Machine Learning for Detection of Malicious Websites on an Unbalanced Dataset. Digital.

All 16 papers, including 12 conference papers: Publications.

Presentations

  • 2026. Project presentation: HPCAI11 ARK: Advanced Machine-Learning Platform for Atomistic Simulations of Reactivity and Kinetics. IKUR-HPC&AI project session, Ikerbasque, CIC nanoGUNE, San Sebastián, 17 July 2026.
  • 2025. Poster: Implementation of a Machine Learning Force Fields Platform for Quantum Dots. AI4AM 2025, Artificial Intelligence for Advanced Materials, San Sebastián, 8–10 April 2025.

Workshops and service

  • Participant, BCMaterials PhD Day, Leioa (28–29 May 2026).
  • Participant, Workshop: New Materials for a Better Life! 2023, BCMaterials, Leioa (26 October 2023).
  • Organiser, 17th International Conference on Emerging Technologies (ICET 2022), GIKI.
  • Instructor, Seminar on Machine Learning with Python, University of Malakand (2022).

Languages

English (fluent), Urdu (native).

↑