About me
👋 My name is Stefanos Laskaridis and I am a Machine Learning Research Scientist specialising in Machine Learning (ML), Distributed & Mobile Systems, and Efficient ML algorithms. My research interests revolve around the areas of efficient LLM deployment, dynamic network architectures (early-exits, MoE, slimmable networks) as well as federated/collaborative learning and on-device AI.
Currently, I am an Applied Scientist at Amazon Science, where I focus on Large Language Models (LLMs) for Alexa+, particularly in the areas of distributed training, post-training Parameter-Efficient Fine-Tuning (PEFT) and Speculative Decoding in multi-turn conversational and agentic tasks.
Previously, I worked as a Machine Learning Researcher at Brave Software and as a Visiting Researcher at the University of Cambridge, where I focused on on-device GenAI, privacy-preserving machine learning as well as modular LLM architectures (Mixture-of-Experts). Before that, I spent 4.5 years as a research scientist at Samsung AI Center in Cambridge, UK, working in the areas of distributed, collaborative and efficient Edge AI. I graduated from the University of Cambridge and, prior to that, I worked at CERN, in Geneva, contributing to large-scale distributed storage systems.
In my free time, I love travelling, motorsport and photography. I am also passionate about the topics of open-source, AGI and privacy and like partaking in hackathons from time to time. Oh, I also like coffee ☕️ …
Recent News
- 01/05/2026 - 📄 Two papers accepted at ICML’26: “FlexRank: Nested Low-Rank Knowledge Decomposition for Adaptive Model Deployment” (spotlight) and “MoSE: Mixture of Slimmable Experts for Efficient and Adaptive Language Models”
- 22/04/2026 - 🎓 Attending ICLR’26 in Rio de Janeiro, Brazil. 🇧🇷
- 21/03/2026 - 🎓 Workshop “Resource-Adaptive Foundation Model Inference” (AdaptFM) accepted at ICML’26 in Seoul, South Korea.