Agenda
- Knowledge graphs as a heterogeneous data store
- Analysis of knowledge graphs
- Graph RAG
- Graph neural networks
- Understanding complex biological mechanisms of action (MoA) for drugs and diseases is crucial for effective ORC prioritization and target identification. MIDAAS, our Multi-modal Integrated Data, Analytics & Agentic System, leverages GenAI-powered multi-agent systems to enhance this biological understanding. However, rigorously evaluating the performance of such systems is essential. This session will introduce MIDAAS and its capabilities in MoA explanation, while also showcasing LEMuR, an inner-source Python tool we developed to streamline the benchmarking of LLMs and multi-agent systems. We will demonstrate how LEMuR addresses the specific evaluation needs of MIDAAS and similar LLM-based systems, ensuring reliable performance.
Speakers
- Prof. Conrad Bessant, Queen Mary University
- Brian Martin, Senior Research Fellow, AbbVie
- Tankred Ott, Senior Data Scientist – AI & Analytics CoE, Novo Nordisk
For details of our full Pharma and Life Science AI/ML Training Program and how to register, please click here.