Postdoc in Harness, Swarm and Memory Engineering Computational Biology

The Laboratory of Causal Systems Immunology combines causal inference, probabilistic AI and large-scale in vivo perturbation experiments to uncover how genes shape immune-cell states and predict how the immune system responds to interventions. This tight integration of advanced machine learning and experimental immunology allows us to tackle fundamental biological questions with direct biomedical relevance. We are embedded in both the VIB Center for AI & Computational Biology—a young, ambitious center whose research groups address biological questions through the lens of AI—and the VIB-UGent Center for Inflammation Research, where we collaborate closely with leading experimental immunologists to unravel the cellular and molecular mechanisms of inflammatory diseases.

The lab is looking for a postdoc to further develop agentic pipelines that autonomously manage computational tasks:

• You oversee dozens of computational biology tasks across biological disciplines
• You launch agent swarms solving these tasks and monitor their progress
• You communicate with task experts to understand ongoing and emerging biological needs and translate them into new tasks
• In collaboration with our wet-lab team and collaborators, you identify and integrate validation data
• In collaboration with the VIB.AI ML unit, you manage a large codebase containing the harness, interpretation, memory, tasks, swarm behaviour and source datasets
• You formulate and test hypotheses that may improve the harness, long-term memory and/or swarming behaviour
• You present your research at international conferences and publish in scientific journals
• You contribute to the supervision of PhD students working in these areas

• Excited to bring biology into the agentic era
• Broad interest across biological domains is a must
• PhD in computational biology, computer engineering, computer science, (bio) statistics, artificial intelligence, physics, or related.
• Desire to push the frontier of both biology and (agent-driven) software engineering
• Fluency in written and spoken English
• Embraces complexity

• An interdisciplinary environment spanning wet-lab, technology-development and computational teams
• Access to unpublished validation data unavailable to existing agents
• State-of-the-art computing capable of running dozens of local large language models concurrently
• Data from emerging technologies through core facilities and tech watch
• Direct connection to a growing team of LLM engineers at VIB
• Part of a dynamic research center with other groups sharing this missionA competitive salary and full benefits