From Theory to Application: Using LLM, RAG, and Embeddings for Real-world Use Case
Join us for an evening of networking and discussions on how new technology can be applied in real-world situations at our AI Happy Hour event! For this session, Julien Baudru and Lluc Bono – IRIDIA, ULB – will explore using new technologies like Large Language Models (LLMs) in real-world scenarios, emphasizing the importance of context for accurate information and outlining steps for smooth integration.
Practical information:
Georganiseerd door:
Description
The presentation explores how new technology can be applied in real-world situations. It discusses the use of Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), and embeddings, and how they can be adapted to improve over time. It highlights the importance of providing the right context to these models to ensure they give accurate information. Key benefits include reducing errors, identifying sources clearly, and improving search capabilities. The presentation outlines a step-by-step process, from gathering data to generating responses, showing how these technologies can be integrated smoothly. Overall, it emphasizes how important it is to consider context when using these advanced tools in various real-life scenarios.
A balanced mix of learning and relaxation – Attendees will have the opportunity to delve into new technologies applications with Julien Baudru and Lluc Bono, followed by a casual happy hour with snacks and beverages!
Teachers / speakers
Julien Baudru
Julien is a PhD candidate at the FARI institute and at the Institut de Recherches Interdisciplinaires et de Développements en Intelligence Artificelle (IRIDIA) of the Université Libre de Bruxelles. After his bachelor’s degree, he completed his master’s in Computer Science with a specialization in artificial intelligence and algorithms at the Université Libre de Bruxelles. His current research focuses on the optimization of multi-modal transport. The idea of multi-modal routing is to adapt the famous shortest path problem to road networks combining several modes of transport. One way of achieving this is to construct the network as a layered structure, where each layer corresponds to a mode of transport. Thus, the ultimate goal of this research will be to find a solution to this mobility problem by including the notions of time, price and ecological impact. The second underlying objective of this research is to enable users of car-sharing solutions to optimize their trips. This is where this project comes into its own for the common good, by designing algorithmic solutions to ultimately reduce the ecological impact of transport, first in Brussels and then, we hope, everywhere.
Articles on carpooling algorithms:
- Julien Baudru & Hugues Bersini - The Comparative Analysis of Car-Pooling Algorithms for Ride-Sharing Systems
- Julien Baudru & Hugues Bersini - Heuristic Optimal Meeting Point Algorithm for Car-Sharing in Large Multimodal Road Networks
Lluc Bono
Lluc is currently a PhD student at Artificial Intelligence research laboratory (IRIDIA) of the Université Libre de Bruxelles and FARI Institute. He finished his bachelor’s in industrial engineering at Universitat Politecnica de Valencia (Spain) in 2016 and his master’s in computer science and Engineering at Université Libre de Bruxelles (Belgium) in 2021. Prior to his masters, he also worked as an Assistant at the Technology Transfer Office of Universitat de les Illes Balears (Spain) where he worked on developing a project on process improvement and KPI automation. At present, he is researching Music and AI, while also being heavily involved in one of FARI’s pilot projects on Animal Welfare and AI.
Related courses
Summer School on Security and Privacy in the Age of AI
Summerschool - Heverlee - KU Leuven, UGent, VUB, Imec