AI300: AI Technology Implementer UPDATE

Agenda:

• Python and data preparation
  

  • Python basics for data work: syntax, data types and functions

  • Working with interactive notebooks and libraries

  • Loading data from CSV and Excel files

  • Exploring, merging and transforming data with Pandas

  • Handling missing values, duplicates and inconsistent data types

  • Creating and interpreting simple visualizations

  • Hands-on lab: Converting raw data into an analyzable and reliable dataset
    
    

• Machine learning in practice

  • Basics and typical use cases of classification and regression

  • The machine learning workflow from use case to evaluation

  • Splitting data into training, validation and test data

  • Preparing data and training models with scikit-learn

  • Implementing reproducible workflows with pipelines

  • Defining baselines and selecting appropriate evaluation metrics

  • Comparing models and interpreting results

  • Detecting and avoiding overfitting and data leakage

  • Hands-on lab: Developing, evaluating and specifically improving a machine learning model
    
    

• Deep learning and generative AI

  • Basic principles of neural networks and transformers

  • Practical application of pre-trained models and transfer learning

  • Using foundation models via an API or model pipeline

  • Basics of prompting and structured model outputs

  • Introduction to embeddings, semantic search and Retrieval-Augmented Generation (RAG)

  • Distinguish between classical ML, RAG, and fine-tuning in a meaningful way

  • Realistically evaluating quality and risks: hallucinations, bias, data privacy, prompt injection, costs and computational requirements

  • Hands-on lab: Designing and implementing a simple end-to-end AI prototype and deriving next steps for your own work environment

Objectives:

Upon completion of the workshop AI300 AI Technology Implementer participants will be able to:

• independently handle basic data tasks with Python and Pandas and understand existing Python code

• import, standardize, clean, and validate data from typical sources

• apply the basic machine learning workflow in practice

• train, compare, and evaluate models using appropriate metrics

• identify typical errors such as overfitting and data leakage

• explain the basic principles of deep learning, transformer and foundation models

• apply pre-trained models and simple RAG approaches in practice

• differentiate when classical ML, RAG or fine-tuning is appropriate

• realistically assess the benefits, data requirements, effort, costs and risks of an AI use case

Target audience:

The training AI300 AI Technology Implementer is targeted at:

• Beginners in Python with a focus on data: data analysts, BI/reporting, (junior) data science, (junior) data engineering

• Departments that want to process/evaluate data themselves (e.g., controlling, operations, marketing analytics)

• Product/project roles that want to better understand, assess, and support AI projects

• Professionals with an interest in technology who are looking for a hands-on introduction to machine learning and generative AI

Prerequisites:

For participation in the course AI300 AI Technology Implementer no specific prerequisites are required.

A basic understanding of data and digital workflows is helpful, but not a prerequisite. All required Python fundamentals are taught compactly and practically in the course. The course is a hands-on training: The participants actively program along on all three days.

Description:

The course AI300 AI Technology Implementer provides provides a comprehensive introduction to working with data and artificial intelligence in practice. The participants prepare data with Python, visualize it, and use it to develop their first machine learning model. They then explore deep learning,foundation models, and retrieval-augmented generation using concrete applications.

The didactic focus lies on hands-on learning: Short technical inputs alternate with guided exercises, interactive notebooks, and realistic mini cases. Readable and reproducible code, clean data, and a responsible handling of AI are central themes throughout all three days. This results in a small portfolio consisting of data analysis, a machine learning pipeline, and a GenAI prototype, which serves as a starting point for participants’ own use cases.