Developing LLM-Enabled Systems 7.5 credits

The course focuses on practical aspects of developing AI systems enabled by LLMs. Students will explore both theoretical foundations and hands-on development techniques, gaining experience in integrating LLMs into various applications, including agentic AI systems capable of autonomous task completion. The course includes the following elements: - Foundation of LLMs - LLM capabilities, limitations, and deployment options, including when using cloud-based vs. self-hosted LLMs. - Prompt Engineering and techniques for optimizing model responses. - Fine-Tuning & adaptation: Customizing LLMs for specific tasks and domains. - Retrieval-Augmented Generation (RAG): Combining LLMs with external data sources. - Testing and quality Assurance: ensuring reliability, safety, and performance of LLM-enabled systems. - Agentic AI: Designing systems where LLMs take initiative, use tools, and act autonomously. - System architecture: designing scalable, efficient, and cost-effective LLM-based applications. - Evaluation & Optimization: performance metrics, latency management, cost trade-offs, and assessing agentic behavior.