01
Foundation Models & Agentic Intelligence
How can foundation models reason, collaborate, and adapt over long-horizon tasks with external knowledge?
Agents · memory · knowledge-guided reasoning
Tuan-Cuong Vuong
AI Researcher · Hanoi
PhD applicant · 2027
Building AI that reasons across language, time, and structured knowledge.
I study how foundation and agentic AI systems can reason over heterogeneous knowledge — language, images, temporal observations, and structured relations. Healthcare and biomedicine are my primary application domains, where reliable reasoning matters most.
Research Assistant · 2021–Present
Applied AI (A2I) Lab · Phenikaa UniversitySupervised by Dr. Trang Xuan Mai
Research arc
Agentic reasoning Structured knowledge Multimodal healthcare
Research agenda
The questions that drive my work
One agenda connects my projects: making intelligent systems reason over heterogeneous evidence, then testing those systems in domains where context, structure, and reliability matter.
02
Multimodal & Structured Reasoning
How can models integrate language, images, temporal signals, and graphs while preserving meaningful relationships?
Multimodal learning · temporal modeling · graphs
03
AI for Healthcare & Biomedicine
How can reasoning systems remain reliable when scientific data are sparse, heterogeneous, and high-stakes?
Clinical prediction · medical imaging · trustworthy AI
Selected research
A coherent trajectory, not a topic list
These projects trace the progression from multi-agent language systems to structured and multimodal reasoning in healthcare.
01 Agentic reasoning
Multi-Agent NLP LLM Published First author
A training-free mixture-of-agents framework for multi-document summarization using LLMs and knowledge graphs
Tuan-Cuong Vuong, Trang Xuan Mai, Tien-Cuong Nguyen, Vu-Duc Ngo, Thien Van Luong
Neural Computing and Applications · 2026
Contribution: As first author, I formulated the research problem, designed and implemented the training-free mixture-of-agents method, and led the experimental evaluation across English and Vietnamese benchmarks.
02 Structured knowledge
Machine Learning Healthcare Knowledge Graphs Multimodal Learning Preprint First author
PIKE: Patient-Context Informed Knowledge Graph Enrichment for ICU Outcome Prediction
Tuan-Cuong Vuong, Trang Xuan Mai, Son Thai Mai, Duong Tung Ta, Vu-Duc Ngo, Tien-Cuong Nguyen, Huan Vu, Thien Van Luong
Preprint · 2026
Contribution: As first author, I formulated the research problem, designed and implemented the patient-context knowledge graph enrichment method, and led the experimental evaluation on sparse, short-sequence ICU records.
03 Multimodal healthcare
Machine Learning Healthcare Multimodal Preprint First author
MCPL: Learning Multi-Component Patient Representations from Multimodal ICU Records for Early Clinical Outcome Prediction
Tuan-Cuong Vuong, Trang Xuan Mai, Tien-Cuong Nguyen, Trong-Nghia Nguyen, Thien Van Luong
Preprint · 2026
Contribution: As first author, I formulated the research problem, designed and implemented the multi-component patient-state representation method, and led the experimental evaluation on multimodal ICU outcome prediction.
Recent updates
Research milestones
Selected publications, acceptances, and research releases.
Jun 2026
Our paper on Training-Free MoA for Multi-Document Summarization was published by Neural Computing and Applications (Q1, Impact Factor: 4.50, SJR: 1.102).
April 2026
Our paper HERMES (Knowledge Graph Reasoning for Patient Outcome Prediction) was accepted at CITA 2026.
PhD opportunities · Research collaboration