Agentivium AI is led by a research lead, advised by faculty, and supported by student researchers working across agent-native systems, multi-agent orchestration, and real-world testbeds.
Founder and Research Lead of Agentivium AI. His work focuses on agent-native systems, multi-agent orchestration, trustworthy autonomous execution, and real-world AI testbeds.
Associate Professor at HCMUT. Founder and director of the High Performance Computing Lab, researching High Performance Computing, Big Data Analytics, Artificial Intelligence, Cloud computing, and network protocols.
Focuses on designing local control memory architectures for Orchestrator Agents in Multi-Agent Systems. His research aims to improve agent coordination by treating memory not just as a shared data store, but as a dynamic control substrate. This memory directly supports the orchestrator in making intelligent scheduling decisions, selecting context, enforcing access controls, and adapting based on execution history across task episodes.
Focuses on runtime optimization for multi-agent execution across distributed clusters. His research addresses how the runtime should intelligently spawn, place, scale, or cancel sub-agents to simultaneously optimize answer quality, latency, and execution cost, while strictly respecting shared-context constraints and reduction needs of agent workflows.
Alumni will be listed here as the research program matures.
