Past Programs
Power Systems and Artificial Intelligence: An Introduction
Power Systems and Artificial Intelligence, led by Program Directors Professor Le Xie and Professor Minlan Yu, introduced learners to the growing intersection of electric power systems, artificial intelligence, and sustainable computing. This in-person short course connected foundational concepts in grid operation with emerging AI methods used for forecasting, monitoring, optimization, renewable energy integration, and grid modernization. It also examined how increasing AI and data center demand is reshaping complex power-system planning and operations.
Learners, such as power systems engineers, device vendors, and data center operators, gained a shared technical vocabulary for working across the power and AI fields, along with practical exposure to statistical time series analysis, dimensionality reduction, computer-based simulations, and energy-efficient AI approaches. Participants emerged better equipped to distinguish between first principles, data-driven, and physics-informed models; evaluate where AI can add value in real-time power-system operations; and consider the reliability, scalability, and environmental impacts of deploying AI within energy infrastructure.
Interested in this topic? Join the Power and AI Interest List to hear about similar future offerings.
Large Language Models: From Transformer Basics to Agentic AI
Large Language Models: From Transformer Basics to Agentic AI, led by Program Director Pavlos Protopapas, introduced learners to the foundations and practical applications of modern large language models, building towards designing intelligent systems to solve real-world problems with confidence and creativity. The course, offered in both in-person workshop and live online formats, covered text preprocessing, neural networks, transformer architectures, and models such as BERT and GPT, before moving into hands-on work with Hugging Face and OpenAI APIs. Learners also explored retrieval augmented generation (RAG), vector databases, agentic workflows, model evaluation, and parameter-efficient fine-tuning techniques such as LoRA.
Learners, from ML engineers to developers to researchers, gained the ability to design more effective prompts, connect LLMs to external systems, build RAG-based applications, fine-tune models for specialized tasks, and develop agentic workflows for automation. The course emphasized practical experience applying LLMs to real-world use cases and honed a stronger understanding of how to integrate and adapt these models responsibly and effectively in technical or organizational settings.
Interested in this topic? Join the Large Language Models Interest List to hear about similar future offerings.
Interested in Learning More?
Explore our current programs or join the SEAS Professional Education interest list to hear about new courses, learning experiences, and future offerings from Harvard SEAS Professional Education.