aiLights - The Potential of Small Language Models in the Development of AI Agent Systems
When people think of modern AI, they often imagine massive cloud-based models with billions of parameters. Yet some promising advances in AI agents (systems that can reason and plan) may come from much smaller language models. These compact models are faster, more energy-efficient, cheaper to deploy, and capable of running locally on everyday devices, making them powerful building blocks for practical AI systems.
This talk explores how small language models can enable modular and scalable AI agents. We will examine the trade-offs between size and capability, the advantages of specialization, and how combining small and large models can create more robust systems. Rather than assuming that bigger is always better, we consider how thoughtful design and efficient models may shape the next generation of intelligent agents.
Speaker:
Prof. Mascha Kurpicz-Briki
Bern University of Applied Sciences, Generative AI Lab & Makubri Technologies Sàrl
Mascha Kurpicz-Briki is an AI expert and professor in computer science. She is the co-lead of the Applied Machine Intelligence research group and the Generative AI Lab at the Bern University of Applied Sciences in Biel, Switzerland, and is working on applied natural language processing (NLP) and machine learning for different use cases.
As keynote speaker, author and founder of Makubri Technologies Sàrl she additionally works on creating formats that make artificial intelligence and digital technologies understandable and engaging.
📅 Date & Time: 30.03.2026, 12:00-12:45 (CET)
⏱️ Duration: 45 minutes
🌐 Language: English
🗣️ Format: Presentation + Q&A
📍 Location: Online (Zoom + YouTube Livestream)
👥 Audience: Broad Public, C-Level, AI Users, IT Architects
🔗 More information about aiLights: https://ailights.org/
We thank our partner Oertli-Stiftung for the support.
If you have any questions, please reach out to:
📧 [email protected]