fam-bluemer.de

What It Is

This site is currently the lecture-notes repository for the short course “From Language Models to AI Agents,” part of the Bachelor Data Science module “Topics in Machine Learning and Data Science” at Katholische Universität Eichstätt-Ingolstadt (KU). The text indicates that the course is scheduled for four Wednesdays between November and December 2025, in room GEOG-101. It looks more like university course material than a commercial online course sold to the public.

Core Content and Teaching Format

The course focuses on artificial intelligence, NLP, large language models, RAG, and AI Agents. It starts with foundational NLP methods such as Bag of Words, N-gram, and Word2vec, then moves on to transformers, decoder-only LLMs, training paradigms, inference parameters, and further topics including RAG, PEFT, multimodal and reasoning models, and agent frameworks such as Model Context Protocol. The format is explicitly described as an in-person short course; there is no indication of livestreaming, recorded videos, or 1-on-1 instruction. The lecture notes are in English, while the actual teaching language is not specified.

Pricing, Certification, and Instructor Background

The page does not disclose pricing, payment methods, or a public enrollment link, nor does it state whether a certificate is provided. Since the course is part of KU’s undergraduate Data Science module, the intended audience is likely mainly on-campus Bachelor Data Science students. The institutional background is clear: it is an official course module at KU in Germany. However, the scraped text does not show the specific instructor’s name or qualifications.

Pros and Cons

The main advantage is that the topics are up to date, connecting traditional NLP with modern LLMs, RAG, and Agentic AI in a structured way. The course also requires students to participate actively, conduct additional research, and complete experiments or small projects, with possible access to resources such as OpenAI API keys and an LLM server. The downside is that the course is very short, and the text explicitly says it does not include prompt engineering or Python coding. For learners who want systematic engineering practice, code-level training, or full project deployment, the depth may be limited. Public information is also missing on grading criteria, assignment details, certification, and fees.

Who It’s For and Access from China

It is suitable for undergraduates or self-learners who already have a machine learning/data science background and want to quickly build a technical map of LLMs. As for access from China, domain connectivity cannot be determined from the page text alone. Related services such as OpenAI, some external platforms, or academic cloud resources may face network or account restrictions in mainland China. Alternatives include DeepLearning.AI, Coursera RAG courses, Hugging Face tutorials, and large-model courses offered by Chinese universities or domestic platforms.