Alexander Acker
Co-founder & CEO at logsight.ai
Dr. Alexander Acker is the co-founder and CEO of logsight.ai, where he leads a team dedicated to enabling companies to efficiently manage their increasingly complex AI model ecosystems. Holding a Ph.D. from TU Berlin and authoring over 30 scientific publications, Alexander has worked with industry giants such as Siemens, Deutsche Telekom, and Huawei and helped to optimize the performance of large-scale data centers, backbone networks, and AI applications. Alexander works on bridging the gap between research and practical implementation, leading to the widespread adoption of innovative technologies. His passion lies in unlocking the potential of efficient operations within complex software systems and large-scale AI models, driving transformative value across different industries.
You can find him and follow his journey on LinkedIn.
speaker topic 2024
Optimizing LLM Operation: Balancing Cost, Efficiency, and Sustainability
The rapid expansion and widespread adoption of foundation models, particularly large language models (LLMs), have introduced significant challenges related to operational efficiency, cost management, and environmental sustainability. As these models grow in size and widespread adaption, the computational demands for both training and inference intensify, leading to escalating financial costs and substantial energy consumption. This talk will address the need to optimize the operations of LLMs, focusing on the latest innovations designed to mitigate these challenges. We will explore strategies for enhancing the scalability and accessibility of LLMs, with particular emphasis on low-resource environments where GPU availability is limited. By exploring key optimization techniques, we will uncover their potential to reduce costs, boost performance, and lower the environmental impact of LLM operations. Furthermore, this session will highlight the critical importance of minimizing the carbon footprint associated with training and deploying these models, underscoring the broader implications for sustainability in AI development. Ultimately, this talk aims to provide a comprehensive overview of the current challenges and state of the art solutions in optimized LLM operation, with a focus on making these models more cost-effective, energy-efficient, and sustainable, empowering SMEs, universities, and individual contributors to play an active role in shaping the future of foundation model development.
speaker topic 2023
At applydata summit 2023 Alexander Acker talked about “Towards the Automation of Data Cleaning and the Role of Clean Data in ML/AI Projects”.