Samuel KANGONI MATIA

Samuel KANGONI MATIA

Transitioning toward AI security and computer vision. Interested in machine unlearning, deepfakes, and AI agentic security. You can download my full academic CV here.

Samuel Kangoni MatiaAI Researcher

I’m Samuel Kangoni Matia, a Computer Engineering graduate from the University of Kinshasa, passionate about solving real-world challenges through Artificial Intelligence.

🔬 Research Interests

Transitioning toward AI security and computer vision. Interested in machine unlearning, deepfakes, and AI agentic security. Passionate about building robust, safe, and trustworthy AI for real-world applications.

Previously worked on:

  • Graph Neural Networks
  • Recommender Systems
  • AI for Healthcare
  • Natural Language Processing

📄 Publications

Journal Articles

  • Samuel Matia Kangoni, Obed Tshimanga Tshipata, Pierre Sedi Nzakuna, Vincenzo Paciello, Jean Gilbert Mbula Mboma, Jean Robert Makulo, Kyandoghere Kyamakya.
    Enhancing Sentiment-driven Recommender Systems with LLM-Based Feature Engineering: A Case Study in Drug Review Analysis.
    IEEE Access, 2025.

    Developed a sentiment-driven drug recommender system leveraging instruction-tuned LLMs for feature engineering. Introduced the Adaptive Confidence-Weighted Scoring (ACWS) mechanism for dynamic weighting and robust outlier handling. Conducted extensive evaluation using standard RecSys metrics and expert clinical validation, demonstrating improved interpretability and recommendation quality. Paper • Code

  • Henri Kasongo Shabani, Kasra Mortazavi, Witesyavwirwa Vianney Kambale, Mahmoud Hamed, Samuel Matia Kangoni, Kyandoghere Kyamakya.
    A Neuro-Symbolic Edge Stack for Fragile Economies – Binary Cellular Neural Networks and Auto-Mined ASP Rules for Robust Econometrics in the Democratic Republic of Congo.
    WSEAS TRANSACTIONS on BUSINESS and ECONOMICS. DOI: 10.37394/23207.2026.23.20

Conference Papers

  • Samuel Matia Kangoni, Obed Tshimanga Tshipata, Jean-Gilbert Mbula Mboma, Pierre Sedi Nzakuna, Vincenzo Paciello, Vianney Witesyavwirwa Kambale, Kyandoghere Kyamakya.
    Structural Preservation in Efficient Pharmacovigilance Summarization: A Causal Graph-Based Evaluation Framework.

    Proposed a causal-graph-oriented evaluation methodology for pharmacovigilance summarization systems. Demonstrated how structural preservation metrics improve safety-critical summarization pipelines in real-world drug surveillance. IEEE AFRICON, 2025.

  • Samuel Matia Kangoni, Mahmoud Hamed, Vianney Witesyavwirwa Kambale, Asha Choudhary, Kyandoghere Kyamakya.
    Large Language Models in Healthcare: Opportunities, Challenges, and Pathways Toward Trustworthy Clinical AI.
    Accepted at AUTSYS-2025 (Springer LNCS).

Book Chapter

  • Samuel Matia Kangoni.
    Machine Learning in Malware Analysis. Chapter 7 in Mastering Malware Development and Analysis: A Comprehensive Guide to Hybrid Malware Analysis and Virtual Sandboxing.
    Apple Academic Press (2026). Publisher page

🧪 Research Experience

Research Mentee – BASIRA Lab (GNNs for Rising Stars)

Imperial College London — Supervisor: Dr. Islem Rekik
Oct 2025 – Present
Selected for a competitive international research mentorship program. Completed with highest distinction. Work on GNN architecture design, explainability, and reproducible ML.

Researcher – BASTION Lab

University of Kinshasa — Supervisor: Dr. Kyandoghere Kyamakya
Sep 2025 – Present
Research on neuro-symbolic AI, Binary Cellular Neural Networks, and Answer Set Programming for econometric modeling and medical imaging.

💼 Professional Experience

  • AI Trainer – EAZYTraining / EASYCONSULTING (France) — Aug 2025 – Present
  • AI Engineer – CongoExcelTechnologies (DRC) — Jul 2024 – Mar 2025
  • AI Engineer Intern – PathOr Platforms (India) — Sep 2024 – Nov 2024
  • AI Consultant – IA Flex Start (France) — Dec 2023 – Mar 2024

🌍 Selective Programs

  • A2SV (Africa to Silicon Valley) – G7 Remote Education Program
    Selected among 166 out of 2,264 applicants (~7% acceptance rate)

  • TReND in Africa – Computational Neuroscience & ML Summer School (Kenya)
    Selected among 20 out of 1,060 applicants (~1.9% acceptance rate) — later declined to focus on AI security & Computer Vision

🏆 Awards & Distinctions

  • 🥇 1st Prize – Hackathon TechPoly 2024 (Mat.IA: AI-powered multilingual tutor)
  • 🥈 2nd Prize – Orange Summer Challenge 2023 (Utulivu: autism screening system)
  • Huawei Seeds for the Future 2024 – Top 10 student representing DRC

📫 Contact

📄 Academic CV

Download my full CV here.

Feel free to reach out for research collaborations.