AI Engineering - Data Science and Machine Learning Engineering Bootcamp

Students are introduced to the overall structure and expectations of the course. They learn how to organize their learning, manage time and stress, and develop a mindset that supports professional growth. An initial overview of the IT industry helps them understand their future work environment, while the first steps in career development and personal branding help set long-term goals. A dedicated session on how to responsibly use AI tools prepares students to integrate AI thoughtfully into their learning. Technical setup ensures everyone is ready to get started.

Students will learn the basic set of skills needed to work in a unix shell. They also recap on general programming Python language fundamentals and write and execute their first small programs. By using Git and Github, they are introduced directly to modern collaborative work in the industry, including code versioning, as well as working collaboratively on the codebase with others through the use of concepts such as branching and pull requests.

Building upon the fundamentals of Phase 2, students are introduced to tools for extracting and manipulating data from different sources, like files but also databases. To do this they use SQL and Pandas. After practicing data manipulation skills they dive into creating descriptive analysis using various plotting libraries and plot types. This phase is completed with a 2 days project focused on Exploratory Data Analysis (EDA) where students work on a real-world dataset. Always with the business case in mind, they provide their artificial stakeholder with recommendations and meaningful visualizations tailored to their needs.

This section covers the basic concepts of supervised machine learning techniques for regression and classification. Students are introduced to multiple algorithms, like linear and logistic regression, decision trees, and ensemble methods, among others. Besides introducing the statistical details of the algorithms, we also consider the scope and assumptions of these models. Students apply all of their learned skills by working in groups on a 4 day machine learning project that covers the whole data science life cycle. Together they set milestones and consider the value of their data product, while gaining experience in collaborative work with Git and Github. The final step is a presentation to the stakeholders.

With a basic understanding of statistical learning techniques and the underlying software implementation, students dive deeper into techniques for forecasting on Time Series. In a world with mostly unlabeled data, it is important to tackle also the challenges of unsupervised learning. Students will get an introduction to both dimensionality reduction and clustering techniques. To get some understanding of concepts that are nowadays often referred to as AI, students get an introduction to artificial neural networks. Building their own deep neural network from scratch will improve their understanding of the underlying processes and concepts. Also typical areas of application like Natural Language Processing and common related techniques like transfer-learning will be addressed.

To condense all the skills learned and also focus even more on group work and collaboration, the capstone project gives the context and time to work on a bigger data science problem from start to end. Students have the chance to find a problem by their own, using publicly available datasets or work with data provided by one of our partner companies. In teams of 3 to 4 people they work towards achieving and presenting a solution to the given problem. Thereby, not only the gained knowledge regarding technical topics and agile methodologies will be applied, but they will also extend their skills based on the requirements of their projects.

In the first phase students will become familiar with software engineering practices and how they relate to data science. The objective of the first week is to write better code when working with data science projects. In order to achieve this we will cover software engineering in Python (writing programs, working with git and object-oriented programming). Then we will show how to bridge the gap between the usual data science workflow and production-ready code.

At the end of this phase students will be comfortable with getting data for their models from many different sources in different formats. Data engineering is about moving and transforming data from one place to another in a reliable and trustworthy way. Students will get introduced to data architecture design for batch and real-time data processing. They will learn how to get data from various sources like database access and APIs. They will then learn the concepts of data modeling with DBT. Following that they will build data pipelines with Prefect and learn the concepts of batch processing and streaming. Finally, they will set up a feature engineering pipeline in the cloud for their Data science project.

In the third phase of the bootcamp the students will get familiar with the machine learning lifecycle and how to bring data science products to production. There will be an introductory session on machine learning basics followed by sessions on testing, deployment strategies, and containerization.

In this phase of the bootcamp students will get familiar with what it means to have machine learning products in production working reliably over time. In the previous phase, they learned how to deploy models; now they will learn how to monitor and maintain them.

By the end of this phase , students will be able to understand, build, evaluate AI systems using modern Large Language Model (LLM) technologies. They will develop foundational knowledge of LLM architectures, embeddings, vector search, and prompt engineering; gain hands-on experience constructing Retrieval-Augmented Generation (RAG) pipelines; perform fine-tuning and evaluation of small models; and design custom agentic systems using frameworks such as LangChain, pydanticAI, and MCP.

In this phase, students will learn to deploy and manage LLM applications. They will build FastAPI services, containerize them with Docker, and deploy AI systems to Google Cloud Run. Students will integrate monitoring and observability to ensure reliability, and they will learn to deploy complete RAG and agent pipelines end-to-end.

In the final phase of the bootcamp, students will take on a comprehensive capstone project that brings together everything they’ve learned. They’ll design, build, deploy, and monitor a complete machine learning system that solves a real-world problem. Working in teams, students will operate as a professional MLE group, using best practices from software engineering, data engineering, machine learning engineering, model monitoring, and LLM development. The bootcamp concludes with a presentation and live demo of their solution to instructors and peers.