Lucas Vittor is Research Associate in the Department of Biomedical Informatics at Harvard Medical School, where he works in the lab of Prof. Marinka Zitnik. His research focuses on artificial intelligence (AI) for medicine and science, with an emphasis on the design of scalable multimodal knowledge graphs, graph-based AI models, and graph-reasoning agents that integrate heterogeneous biomedical data. He develops and trains foundation models, custom embeddings, and multi-agent architectures to accelerate hypothesis generation, therapeutic discovery, and precision health applications.
At Harvard, Vittor leads the development of advanced biomedical AI infrastructure, designing and deploying model pipelines for large language and foundation models using multi-GPU training and inference. His work includes building large-scale data pipelines optimized for precision and throughput, training and orchestrating graph-based and multi-agent AI systems, and implementing experimentation platforms that enable scientific discovery across diverse biomedical modalities. He collaborates closely with graduate students and postdoctoral researchers to design transformative AI tools that unify graph-based reasoning with real-world clinical and biomedical data.
Before joining HMS, Vittor was an early data engineer at Mutt Data, where he helped scale the company from its earliest stage into a hundred-engineer organization while leading the product strategy for large-scale machine-learning systems. In that role, he built cost-efficient ML architectures for industry leaders and designed ad-tech platforms managing multi-million dollar budgets.
He holds a B.S. in Computer Science and an M.S. in Information Engineering from the Instituto Tecnologico de Buenos Aires, graduating with an Honor Diploma in the top 1% of his class.